TextPecker
Collection
Rewarding Structural Anomaly Quantification for Enhancing Visual Text Rendering
•
7 items
•
Updated
•
1
id
stringlengths 11
104
| images
images listlengths 1
1
| conversations
listlengths 2
2
| data_source
stringlengths 37
86
| class
stringlengths 37
86
| ori_bbox
listlengths 4
4
⌀ |
|---|---|---|---|---|---|
box_se_qa_451236
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[264, 357, 726, 388]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[264, 357, 726, 388]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"一审法院没有将” 超清，4K，电影级构图\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_cn_7k/image_59855.png
|
Qwen_Image/generated_images_cn_7k/image_59855.png
|
[
"350.54060324825986",
"473.51353441608666",
"963.5406032482599",
"514.5135344160866"
] |
|
/mnt/bn/ocr-generation-lf/zhuhanshen/wuxc/ms-swift/data_our/JREC_EVALFIX_1.06k.json_annotrain_box_355276
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[54, 712, 679, 799]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[54, 712, 679, 799]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Copyright ©\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/6011.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/6011.jpg
|
[
"41",
"617",
"513",
"693"
] |
|
box_se_qa_22220
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[615, 355, 747, 381]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[615, 355, 747, 381]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"HANUKKAH\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images_atlas/image_453218.png
|
SeedDream3.0/gen_images_atlas/image_453218.png
|
[
"629",
"363",
"764",
"390"
] |
|
syn_annotrain_box_333460
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[58, 657, 355, 772]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[58, 657, 355, 772]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"洲<#>/万事达\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/17135.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/17135.jpg
|
[
"58",
"404",
"357",
"475"
] |
|
syn_annotrain_box_201208
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[103, 425, 585, 554]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[103, 425, 585, 554]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"产品类别:\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/6611.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/6611.jpg
|
[
"86",
"321",
"491",
"418"
] |
|
box_se_qa_402296
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[20, 619, 946, 831]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[20, 619, 946, 831]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"朝向<#><#><#>1\"}\n```",
"role": "assistant"
}
] |
CogView4/generated_images_easy/image_6216.png
|
CogView4/generated_images_easy/image_6216.png
|
[
"20",
"633",
"968",
"850"
] |
|
img_se_qa_40386
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"让我们开始吧.\"}\n```",
"role": "assistant"
}
] |
Kolors/generated_images_easy/image_12713.png
|
Kolors/generated_images_easy/image_12713.png
| null |
|
syn_annotrain_img_13345
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"古<#>作<#>-002\\n◆<#><#>:<#>100g\\narbor Intl\\n<#>在亚<#>. <#>便\\n她的<#><#>。在<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/38004.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/38004.jpg
| null |
|
syn_annotrain_box_339974
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[109, 356, 931, 480]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[109, 356, 931, 480]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"3<#>级酒店 (60)\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/19423.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/19423.jpg
|
[
"79",
"309",
"677",
"416"
] |
|
box_se_qa_158625
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[833, 964, 1000, 991]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[833, 964, 1000, 991]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Cnc<#><#>d <#>nc lmeo<#>\"}\n```",
"role": "assistant"
}
] |
SDv15/generated_images_Atlas_20k/image_11431.png
|
SDv15/generated_images_Atlas_20k/image_11431.png
|
[
"852.2804561200924",
"985.8143042725173",
"1023.7354214780601",
"1013.5563654734411"
] |
|
syn_annotrain_box_446529
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[653, 117, 939, 206]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[653, 117, 939, 206]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#><#>号码\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/7285.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/7285.jpg
|
[
"566",
"82",
"814",
"144"
] |
|
box_se_qa_134249
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[962, 902, 996, 922]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[962, 902, 996, 922]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<###>\"}\n```",
"role": "assistant"
}
] |
SDv15/generated_images_Atlas_20k/image_9032.png
|
SDv15/generated_images_Atlas_20k/image_9032.png
|
[
"984.544237012987",
"923.0473214285714",
"1018.4387175324675",
"942.9090909090909"
] |
|
box_se_qa_151831
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[230, 795, 781, 857]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[230, 795, 781, 857]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"colornf offplemini duterfy.\\\"\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_10081.png
|
Flux/generated_images_Atlas_20k/image_10081.png
|
[
"235",
"813",
"799",
"877"
] |
|
syn_annotrain_box_250102
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[20, 277, 840, 386]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[20, 277, 840, 386]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Costa Brava机<#>。\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/35614.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/35614.jpg
|
[
"18",
"240",
"752",
"335"
] |
|
box_se_qa_63539
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[795, 540, 878, 585]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[795, 540, 878, 585]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"三¯\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/number_data/Untitle_20250609_144650_4/765/1.png
|
SeedDream3.0/number_data/Untitle_20250609_144650_4/765/1.png
|
[
"813.5384615384615",
"552.6153846153846",
"898.4615384615385",
"598.1538461538462"
] |
|
syn_annotrain_img_28580
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"0公里,出口货物\\n<#>。\\n东雄营销\\n<#> 查看全部运动<#> \\n<#>司 客服电<#>： \\n– 美国\\n0-6610-01\\n备及其<#><#>件业\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/27058.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/27058.jpg
| null |
|
box_se_qa_180060
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[458, 203, 546, 256]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[458, 203, 546, 256]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"O\"}\n```",
"role": "assistant"
}
] |
AnyText/generated_images_LeX_10k/image_3739.png
|
AnyText/generated_images_LeX_10k/image_3739.png
|
[
"234.0671106557377",
"103.7855862119438",
"279.0319818501171",
"130.76450892857142"
] |
|
box_se_qa_440356
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[357, 325, 646, 359]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[357, 325, 646, 359]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"超清，4K，电<#>级构图\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_cn_7k/image_164582.png
|
Qwen_Image/generated_images_cn_7k/image_164582.png
|
[
"473.99999999999994",
"431.00000000000006",
"857",
"477"
] |
|
esyn_annotrain_box_101931
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[744, 318, 927, 464]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[744, 318, 927, 464]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"R）\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/1/13178.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/1/13178.jpg
|
[
"708",
"231",
"882",
"337"
] |
|
box_se_qa_291776
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[15, 923, 200, 968]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[15, 923, 200, 968]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"21:53:28\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_16427.png
|
Flux/generated_images_Atlas_20k/image_16427.png
|
[
"15.587785421545668",
"944.0625",
"205.03981264637002",
"990.4262295081967"
] |
|
box_se_qa_30790
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[199, 597, 948, 675]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[199, 597, 948, 675]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"他表示，发动袭击的！<#><#><#><#>\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/Untitle_20250501_013833/210/1.png
|
SeedDream3.0/Untitle_20250501_013833/210/1.png
|
[
"204",
"610.8120649651972",
"970.1020881670534",
"691"
] |
|
esyn_annotrain_box_62066
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[67, 355, 893, 491]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[67, 355, 893, 491]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"登录 注册 忘记密码\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/2826.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/2826.jpg
|
[
"58",
"318",
"774",
"439"
] |
|
syn_annotrain_img_27988
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \" 网<#>（英语）：\\n<#>一切<#><#>饮\\n路歌集团\\npo\\n<#>门<#>绳\\n专科机构\\n票\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/25074.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/25074.jpg
| null |
|
syn_annotrain_box_67242
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[576, 821, 675, 911]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[576, 821, 675, 911]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>。\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/4/41467.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/4/41467.jpg
|
[
"483",
"620",
"566",
"688"
] |
|
box_se_qa_139317
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[186, 342, 719, 381]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[186, 342, 719, 381]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Welcome to our pub transortatio<#> center\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_15909.png
|
SD3/generated_images_Atlas_20k/image_15909.png
|
[
"190",
"350",
"736",
"390"
] |
|
box_se_qa_349375
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[57, 403, 976, 645]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[57, 403, 976, 645]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>的发展\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/Untitle_20250504_012408/1081/1.png
|
SeedDream3.0/Untitle_20250504_012408/1081/1.png
|
[
"58.483864163678554",
"412.0700274203755",
"998.9200590592702",
"660.2370807846446"
] |
|
img_se_qa_50689
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"WS\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_14598.png
|
Flux/generated_images_Atlas_20k/image_14598.png
| null |
|
box_se_qa_462515
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[21, 893, 209, 932]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[21, 893, 209, 932]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"habitable planet.\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_Atlas_3k/image_20836.png
|
Qwen_Image/generated_images_Atlas_3k/image_20836.png
|
[
"27.481438515081205",
"1185.0742459396752",
"277.3329466357308",
"1236.2587006960557"
] |
|
syn_annotrain_box_153740
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[29, 414, 481, 501]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[29, 414, 481, 501]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"重置您的浏览<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/26738.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/26738.jpg
|
[
"25",
"301",
"417",
"364"
] |
|
box_se_qa_403764
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[24, 716, 936, 780]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[24, 716, 936, 780]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Agriculture in Indonesia' Prazepmeed —\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_5343.png
|
Flux/generated_images_Atlas_20k/image_5343.png
|
[
"24.743494423791823",
"732.1536942379182",
"957.3828996282527",
"798.1363266728624"
] |
|
box_se_qa_472842
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[84, 220, 319, 283]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[84, 220, 319, 283]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"背景技术。\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_cn_7k/image_191406.png
|
Qwen_Image/generated_images_cn_7k/image_191406.png
|
[
"112",
"292",
"423",
"375"
] |
|
box_se_qa_38518
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[632, 828, 942, 852]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[632, 828, 942, 852]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"the text: \\\"New Study Reveals\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images_atlas/image_160415.png
|
SeedDream3.0/gen_images_atlas/image_160415.png
|
[
"646.5384615384627",
"846.8537342583957",
"964.0532544378704",
"871.2205981637209"
] |
|
syn_annotrain_box_111569
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[34, 67, 829, 247]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[34, 67, 829, 247]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"员的家庭成员或者监护人\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/4/49017.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/4/49017.jpg
|
[
"30",
"56",
"742",
"207"
] |
|
syn_annotrain_box_15308
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[259, 911, 294, 951]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[259, 911, 294, 951]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"。\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/15829.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/15829.jpg
|
[
"239",
"713",
"271",
"745"
] |
|
syn_annotrain_img_68772
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"办公<#>。 <#>隆坡辉\\n妈<#>6647\\n的帮<#>。\\n每天均有一定\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/8478.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/8478.jpg
| null |
|
syn_annotrain_box_413020
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[32, 914, 73, 940]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[32, 914, 73, 940]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"。\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/39841.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/39841.jpg
|
[
"23",
"818",
"53",
"841"
] |
|
syn_annotrain_box_399910
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[123, 398, 592, 471]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[123, 398, 592, 471]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"户<#> 56591\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/28090.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/28090.jpg
|
[
"86",
"345",
"414",
"408"
] |
|
syn_annotrain_box_95234
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[23, 470, 428, 576]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[23, 470, 428, 576]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"观了联邦政府\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/16663.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/16663.jpg
|
[
"17",
"381",
"323",
"467"
] |
|
syn_annotrain_box_130406
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[25, 517, 741, 746]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[25, 517, 741, 746]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>览人次: 1596\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/12198.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/12198.jpg
|
[
"21",
"390",
"622",
"563"
] |
|
img_se_qa_72
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"The drrought in the United'Statees'stinal Parks\\nDr<#>ogus' the National Parks\\nNational Causes\\nD App\\nEcoysstem Disrupption\\nNew App Alllows Userss to t<#>aking personal finances Ea<#>silly.\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_18014.png
|
SD3/generated_images_Atlas_20k/image_18014.png
| null |
|
syn_annotrain_box_55850
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[55, 536, 305, 629]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[55, 536, 305, 629]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"飞行 (143)\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/22828.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/22828.jpg
|
[
"49",
"390",
"273",
"457"
] |
|
box_se_qa_68995
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[22, 620, 163, 658]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[22, 620, 163, 658]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<###>\"}\n```",
"role": "assistant"
}
] |
SDv15/generated_images_Atlas_20k/image_2471.png
|
SDv15/generated_images_Atlas_20k/image_2471.png
|
[
"22.724298007246375",
"633.8309178743962",
"166.49245169082127",
"673.6376811594203"
] |
|
box_se_qa_416355
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[491, 44, 523, 58]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[491, 44, 523, 58]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#><#>eL.\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_15820.png
|
SD3/generated_images_Atlas_20k/image_15820.png
|
[
"502.3012146854102",
"45.06828624721351",
"534.9476365952412",
"58.93171375278649"
] |
|
box_se_qa_302270
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[702, 470, 951, 625]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[702, 470, 951, 625]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>EE<#><#>A<#>N\"}\n```",
"role": "assistant"
}
] |
AnyText/generated_images_LeX_10k/image_9746.png
|
AnyText/generated_images_LeX_10k/image_9746.png
|
[
"358.75638051044086",
"239.96287703016242",
"485.86542923433876",
"319.55452436194895"
] |
|
syn_annotrain_box_262002
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[87, 77, 674, 167]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[87, 77, 674, 167]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>装；<#><#><#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/25967.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/25967.jpg
|
[
"68",
"65",
"528",
"140"
] |
|
syn_annotrain_box_124692
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[96, 661, 534, 796]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[96, 661, 534, 796]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"er 5<#>助\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/22734.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/22734.jpg
|
[
"91",
"425",
"508",
"512"
] |
|
syn_annotrain_img_72219
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"TY4 东京 IBX® 数<#><#><#>\\nNPA,NPH<#>\\n构（当地<#>可会计<#>事<#>所及法<#>\\n <#>克共和国\\n工控企业\\n内部控制体系建<#><#><#>本方针\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/20659.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/2/20659.jpg
| null |
|
syn_annotrain_box_215581
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[106, 22, 839, 127]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[106, 22, 839, 127]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>局官<#><#>站\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/17689.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/1/17689.jpg
|
[
"71",
"20",
"563",
"117"
] |
|
img_se_qa_69181
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#><#>中<#><#>的<#><#><#>\\n<#><#>学的<#>推推\"}\n```",
"role": "assistant"
}
] |
Kolors/generated_images_easy/image_21193.png
|
Kolors/generated_images_easy/image_21193.png
| null |
|
esyn_annotrain_box_146312
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[301, 417, 692, 557]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[301, 417, 692, 557]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"第2阶段\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/3/37220.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/3/37220.jpg
|
[
"219",
"350",
"503",
"467"
] |
|
esyn_annotrain_box_59644
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[61, 159, 866, 239]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[61, 159, 866, 239]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"送和洗衣房服务。\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/1/13477.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/1/13477.jpg
|
[
"48",
"160",
"678",
"241"
] |
|
box_se_qa_39338
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[681, 573, 702, 584]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[681, 573, 702, 584]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<###>\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images_atlas/image_444465.png
|
SeedDream3.0/gen_images_atlas/image_444465.png
|
[
"696.6540023201857",
"585.7045751160093",
"717.7728393271461",
"596.9239776682134"
] |
|
syn_annotrain_img_5740
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \" Suidobashi\\n方，以便从底部获\\n唤醒服务, \\n・洗脸毛巾\\n今日超值好价 15%\\n共有4条信息,1页 1\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/2/27697.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/2/27697.jpg
| null |
|
syn_annotrain_img_64404
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#> <#>私条款\\n<#>。\\n苏里宾馆 Jeju\\n© 2019 www.por\\n<#><#><#><#> (6)\\ny\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/4057.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/4057.jpg
| null |
|
box_se_qa_16584
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[75, 538, 597, 600]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[75, 538, 597, 600]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \":5S<#><#>0<#>.<#>am<#><#><#><#>\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images/image_259953.png
|
SeedDream3.0/gen_images/image_259953.png
|
[
"76.44728535353535",
"549.9999061561562",
"610.4472853535353",
"613.9999061561562"
] |
|
esyn_annotrain_box_103668
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[173, 66, 571, 156]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[173, 66, 571, 156]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"型录下载 ⅴ\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/2/25177.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/2/25177.jpg
|
[
"140",
"50",
"463",
"118"
] |
|
syn_annotrain_box_437317
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[90, 548, 461, 699]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[90, 548, 461, 699]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"至23:00\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/3191.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/3191.jpg
|
[
"88",
"337",
"451",
"430"
] |
|
syn_annotrain_box_253917
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[40, 462, 542, 573]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[40, 462, 542, 573]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \" <#>一条 为<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/1750.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/0/1750.jpg
|
[
"27",
"426",
"364",
"529"
] |
|
box_se_qa_427620
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[568, 457, 795, 476]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[568, 457, 795, 476]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"As we stnive to reate a mo<#> a and\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_8396.png
|
SD3/generated_images_Atlas_20k/image_8396.png
|
[
"581",
"468",
"813",
"487"
] |
|
box_se_qa_35895
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[579, 758, 633, 788]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[579, 758, 633, 788]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"u<#>\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images_atlas/image_384857.png
|
SeedDream3.0/gen_images_atlas/image_384857.png
|
[
"591.9194460556845",
"775.7734918793503",
"647.6609628770303",
"806.3856583526683"
] |
|
syn_annotrain_img_18296
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \" 3203n）,特征是半径底槽，公制\\n家电等五金冲压件及\\n命更长；配备充电及\\n科列利奇区\\n订阅西门子故事\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/1363.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/1363.jpg
| null |
|
esyn_annotrain_box_45872
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[15, 390, 387, 533]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[15, 390, 387, 533]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"海滩性爱\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/3/33651.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/3/33651.jpg
|
[
"15",
"251",
"379",
"343"
] |
|
img_se_qa_107614
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"'As Siight pours!\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_Atlas_3k/image_237881.png
|
Qwen_Image/generated_images_Atlas_3k/image_237881.png
| null |
|
syn_annotrain_img_34511
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"测<#>及相<#>流程。\\n7/5 星 （\\n上提供的<#><#>提供\\n交易所资讯\\n站需5分<#>。\\nved.\\n<#>区主<#>：常凤<#>医<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/28235.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/28235.jpg
| null |
|
syn_annotrain_img_12284
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"当前位置：首页品牌品牌集合\\n尊敬的用户：\\n向上取得了很\\n例化的类以及必\\n钻石赞助商\\n品，住宿，一整套文件，允许\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/2238.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/2238.jpg
| null |
|
box_se_qa_444028
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[172, 272, 867, 434]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[172, 272, 867, 434]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>柶<#>卡孔\"}\n```",
"role": "assistant"
}
] |
Qwen_Image/generated_images_cn_7k/image_157994.png
|
Qwen_Image/generated_images_cn_7k/image_157994.png
|
[
"228.23665893271468",
"361",
"1150",
"576"
] |
|
syn_annotrain_box_241182
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[25, 237, 620, 323]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[25, 237, 620, 323]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"③ 信息主体输<#><#><#><#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/26823.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/2/26823.jpg
|
[
"25",
"172",
"624",
"235"
] |
|
syn_annotrain_box_95675
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[21, 217, 752, 316]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[21, 217, 752, 316]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"实惠<#><#> (74)\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/30847.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/30847.jpg
|
[
"15",
"182",
"547",
"265"
] |
|
esyn_annotrain_box_147327
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[18, 220, 549, 382]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[18, 220, 549, 382]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"住宿加早餐旅馆在犹他州 (39)\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/1/18527.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/1/18527.jpg
|
[
"17",
"154",
"522",
"267"
] |
|
esyn_annotrain_box_142271
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[229, 29, 876, 123]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[229, 29, 876, 123]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"唱诗班歌手登录\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/3064.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/0/3064.jpg
|
[
"147",
"28",
"563",
"117"
] |
|
syn_annotrain_box_8726
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[147, 458, 727, 542]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[147, 458, 727, 542]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"符合条件<#>原\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/30569.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/30569.jpg
|
[
"123",
"423",
"610",
"500"
] |
|
syn_annotrain_box_435938
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[53, 531, 879, 816]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[53, 531, 879, 816]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"厘米<#>空气一起<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/9983.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/0/9983.jpg
|
[
"52",
"431",
"861",
"662"
] |
|
syn_annotrain_box_203118
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[18, 72, 787, 142]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[18, 72, 787, 142]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"047 10 1 418 340 \"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/35737.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/35737.jpg
|
[
"16",
"56",
"682",
"111"
] |
|
syn_annotrain_img_70831
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"律商<#>》（China \\n资<#>中心\\n<#><#>简介\\n营业<#>照\\ndeway In\\n<#>清<#>片\\n本·<#>策尔\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/4/49775.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/4/49775.jpg
| null |
|
box_se_qa_58072
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[146, 504, 854, 600]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[146, 504, 854, 600]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Global Refugee Crisis\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_539.png
|
Flux/generated_images_Atlas_20k/image_539.png
|
[
"149.67981438515082",
"515.5638051044084",
"873.1322505800464",
"614.1624129930394"
] |
|
esyn_annotrain_box_25547
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[128, 219, 662, 330]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[128, 219, 662, 330]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"纽约时代广场威斯丁\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/3/34174.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/3/34174.jpg
|
[
"122",
"141",
"630",
"212"
] |
|
esyn_annotrain_box_21723
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[141, 327, 704, 431]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[141, 327, 704, 431]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"H106 好驻旅馆\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/0/1573.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_p2/images/0/1573.jpg
|
[
"130",
"274",
"650",
"362"
] |
|
box_se_qa_141263
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[201, 375, 798, 428]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[201, 375, 798, 428]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"information, updates, and\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_8686.png
|
Flux/generated_images_Atlas_20k/image_8686.png
|
[
"205.16914335664336",
"383.31552301864804",
"816.824155011655",
"438.01635343822846"
] |
|
box_se_qa_6121
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[630, 455, 822, 506]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[630, 455, 822, 506]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"云闪银付\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images/image_285883.png
|
SeedDream3.0/gen_images/image_285883.png
|
[
"644",
"465",
"841",
"518"
] |
|
syn_annotrain_box_414622
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[69, 650, 388, 793]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[69, 650, 388, 793]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>房信息\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/11026.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/11026.jpg
|
[
"66",
"418",
"369",
"510"
] |
|
box_se_qa_147755
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[234, 353, 408, 387]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[234, 353, 408, 387]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Welcome to\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_Atlas_20k/image_8164.png
|
Flux/generated_images_Atlas_20k/image_8164.png
|
[
"239.80403451492538",
"360.98262593283584",
"417.53031716417905",
"395.47985074626865"
] |
|
syn_annotrain_box_201651
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[140, 616, 772, 730]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[140, 616, 772, 730]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#><#><#>时间<#><#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/38730.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi_p2/images/3/38730.jpg
|
[
"94",
"603",
"518",
"715"
] |
|
img_se_qa_58721
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"S3入<#>样<#>定定模\\\",,\"}\n```",
"role": "assistant"
}
] |
Kolors/generated_images_easy/image_8421.png
|
Kolors/generated_images_easy/image_8421.png
| null |
|
img_se_qa_38782
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the entire image.\n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"All text in the image (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"延边，坏体,\\n<#>面\\n<#>体右<#><#><#><#><#>延边\"}\n```",
"role": "assistant"
}
] |
Kolors/generated_images_easy/image_5955.png
|
Kolors/generated_images_easy/image_5955.png
| null |
|
box_se_qa_307319
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[651, 777, 997, 804]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[651, 777, 997, 804]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"lin Ho<#>pitia<#>ls Across 00i<#><#>\"}\n```",
"role": "assistant"
}
] |
SDv15/generated_images_Atlas_20k/image_9558.png
|
SDv15/generated_images_Atlas_20k/image_9558.png
|
[
"666.4430585440202",
"794.6678041509772",
"1019.4356284740902",
"822.6685325892056"
] |
|
box_se_qa_295079
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[36, 408, 270, 530]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[36, 408, 270, 530]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"DNDER\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_LeX_10K/image_7627.png
|
Flux/generated_images_LeX_10K/image_7627.png
|
[
"37",
"417",
"276",
"542"
] |
|
box_se_qa_75631
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[172, 419, 353, 447]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[172, 419, 353, 447]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"im<#>enldedvellop<#>ent\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_9709.png
|
SD3/generated_images_Atlas_20k/image_9709.png
|
[
"176.45689809030875",
"428.2844904515438",
"361.45689809030875",
"457.2844904515438"
] |
|
box_se_qa_301727
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[679, 446, 972, 621]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[679, 446, 972, 621]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<###>\"}\n```",
"role": "assistant"
}
] |
AnyText/generated_images_LeX_10k/image_6573.png
|
AnyText/generated_images_LeX_10k/image_6573.png
|
[
"346.877030162413",
"228.08352668213456",
"496.5568445475638",
"317.17865429234337"
] |
|
box_se_qa_157015
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[252, 245, 768, 378]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[252, 245, 768, 378]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"MARINE\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_LeX_10K/image_1530.png
|
Flux/generated_images_LeX_10K/image_1530.png
|
[
"258",
"251",
"786",
"387"
] |
|
box_se_qa_26193
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[99, 394, 960, 527]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[99, 394, 960, 527]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"毛宜全参加的合一\"}\n```",
"role": "assistant"
}
] |
SeedDream3.0/gen_images/image_80169.png
|
SeedDream3.0/gen_images/image_80169.png
|
[
"101",
"403",
"982",
"539"
] |
|
box_se_qa_335816
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[66, 323, 502, 435]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[66, 323, 502, 435]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"企业庆<#>、\"}\n```",
"role": "assistant"
}
] |
Kolors/generated_images_easy/image_17652.png
|
Kolors/generated_images_easy/image_17652.png
|
[
"68",
"330",
"514",
"445"
] |
|
box_se_qa_233102
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[485, 545, 604, 635]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[485, 545, 604, 635]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"DA..\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_LeX_10K/image_7702.png
|
Flux/generated_images_LeX_10K/image_7702.png
|
[
"496.5568445475638",
"557.1415313225058",
"617.7262180974478",
"649.800464037123"
] |
|
box_se_qa_127911
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[68, 459, 894, 652]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[68, 459, 894, 652]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#><#>议<#><#>\"}\n```",
"role": "assistant"
}
] |
CogView4/generated_images_easy/image_21138.png
|
CogView4/generated_images_easy/image_21138.png
|
[
"70",
"470",
"915",
"667"
] |
|
esyn_annotrain_box_103847
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[68, 191, 631, 381]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[68, 191, 631, 381]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"eNGN技术\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/2/23942.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/2/23942.jpg
|
[
"57",
"139",
"529",
"277"
] |
|
syn_annotrain_box_192330
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[18, 582, 317, 675]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[18, 582, 317, 675]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"<#>史出<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/36112.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/3/36112.jpg
|
[
"15",
"537",
"266",
"623"
] |
|
box_se_qa_280646
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[26, 832, 950, 885]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[26, 832, 950, 885]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"Volcanic Ash Cllouds Interefer with\"}\n```",
"role": "assistant"
}
] |
SD3/generated_images_Atlas_20k/image_8192.png
|
SD3/generated_images_Atlas_20k/image_8192.png
|
[
"26.256410256410255",
"850.946386946387",
"971.4871794871794",
"905.8461538461538"
] |
|
box_se_qa_79056
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[616, 147, 986, 405]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[616, 147, 986, 405]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"DO,,\"}\n```",
"role": "assistant"
}
] |
AnyText/generated_images_LeX_10k/image_6783.png
|
AnyText/generated_images_LeX_10k/image_6783.png
|
[
"315",
"75",
"504",
"207"
] |
|
box_se_qa_237828
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[332, 124, 668, 277]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[332, 124, 668, 277]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"2015\"}\n```",
"role": "assistant"
}
] |
Flux/generated_images_LeX_10K/image_6503.png
|
Flux/generated_images_LeX_10K/image_6503.png
|
[
"340",
"127",
"683",
"283"
] |
|
syn_annotrain_box_404820
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[147, 465, 715, 547]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[147, 465, 715, 547]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"150W UFO工矿<#>\"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/15663.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext_zaozi/images/1/15663.jpg
|
[
"132",
"403",
"640",
"474"
] |
|
esyn_annotrain_box_91781
|
[
{
"content": "<image>\n\n This is a text-generated image. Please recognize all visible text in the local area \"bbox_2d:[452, 394, 808, 549]\". \n Marking rules:\n 1. Use <#> for structurally flawed (e.g., extra/missing strokes, distortion) unrecognizable Chinese characters or single English letters;\n 2. Use <###> exclusively for structurally flawed unrecognizable single English words (not multi-word phrases, lines, or sentences).\n Output in the following JSON format:\n {\n \"recognized_text\": \"Text in \"bbox_2d:[452, 394, 808, 549]\" (including structural error markers)\"\n }",
"role": "user"
},
{
"content": "```json\n{\"recognized_text\": \"小+对+单 \"}\n```",
"role": "assistant"
}
] |
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/4/48267.jpg
|
TextEvaluator/hanzi_writer/synthtiger/syth_multitext/images/4/48267.jpg
|
[
"481",
"231",
"859",
"322"
] |
End of preview. Expand
in Data Studio
TextPecker-1.5M: A Dataset for Training and evaluating TextPecker
This repository contains the TextPecker-1.5M dataset, a new benchmark proposed in the paper "TextPecker: Rewarding Structural Anomaly Quantification for Enhancing Visual Text Rendering".
Code and Project Page
The official implementation and project details for the TextPecker and TextPecker-1.5M dataset can be found on the GitHub repository: https://github.com/CIawevy/TextPecker
Sample Usage
You can easily load the TextPecker-1.5M dataset using the Hugging Face datasets library. The dataset is provided in two configurations: train and test
from datasets import load_dataset
# Load the full TextPecker-1.5M dataset (includes train and test splits)
dataset = load_dataset("CIawevy/TextPecker-1.5M", "default")
train_data = dataset["train"]
test_data = dataset["test"]
# Load specific split directly (more efficient for practical usage)
train_data = load_dataset("CIawevy/TextPecker-1.5M", "default", split="train")
test_data = load_dataset("CIawevy/TextPecker-1.5M", "default", split="test")
For detailed instructions on installation, model download, evaluation, and running demos with the FreeFine framework, please refer to the GitHub repository.
Citation
If you find this dataset useful for your research, please cite the accompanying paper:
@article{zhu2026TextPecker,
title = {TextPecker: Rewarding Structural Anomaly Quantification for Enhancing Visual Text Rendering},
author = {Zhu, Hanshen and Liu, Yuliang and Wu, Xuecheng and Wang, An-Lan and Feng, Hao and Yang, Dingkang and Feng, Chao and Huang, Can and Tang, Jingqun and Bai, Xiang},
journal = {arXiv preprint arXiv:xxxxx},
year = {2026}
}
- Downloads last month
- 295
Size of downloaded dataset files:
985 GB
Size of the auto-converted Parquet files:
985 GB
Number of rows:
1,483,089