| --- |
| license: cc-by-nc-sa-4.0 |
| --- |
| # 50000-Sets-Image-Editing-Data |
|
|
| ## Description |
| This dataset is just a sample of 50,000 Sets - Image Editing Data. The editing types include human attribute editing, image semantic editing, and image structure editing. The editing targets cover scenes such as people, animals, goods, plants, and landscapes. In terms of annotation, based on the editing instructions, the targets that need to be edited in the image are edited. The data can be used for tasks such as image synthesis, data augmentation, and virtual scene generation. |
|
|
| For more details, please refer to the link: https://www.nexdata.ai/datasets/llm/1785?source=Huggingface |
|
|
| ## Specifications |
| ### Data size |
|
|
| 50,000 sets |
|
|
| ### Object types |
|
|
| people, animals, products, plants, landscapes, etc. |
|
|
| ### Editing types |
|
|
| human attribute editing, image semantic editing, image structural editing |
|
|
| ### Resolution |
|
|
| no less than 1080p in principle |
|
|
| ### Data parameters |
|
|
| image formats include .jpg, .jpeg, .png, and other common formats; editing text format is .txt |
|
|
| ### Annotation content |
|
|
| edit the object s in the image according to the editing instructions |
|
|
| ### Accuracy rate |
|
|
| the edited image data must comply with the data requirements, with no significant mismatches with the original image; the accuracy rate must not be lower than 95%. The mask edges should be within a 5- |
|
|
| ## Full Dataset Access |
| The complete dataset is available upon request. Visit [Nexdata](https://www.nexdata.ai/datasets/llm/1785?source=Huggingface) to learn more and submit an access request. |
|
|
