Datasets:
Modalities:
Text
Formats:
json
Languages:
English
Size:
< 1K
Tags:
text-to-image
multimodal
indoor-scenes
prompt-engineering
stable-diffusion
scene-understanding
DOI:
License:
update readme
Browse files
README.md
CHANGED
|
@@ -72,7 +72,9 @@ Prompt2SceneBench can be directly used for:
|
|
| 72 |
## Dataset Structure
|
| 73 |
|
| 74 |
### CSV Format (`prompt2scene_prompts_final.csv`)
|
|
|
|
| 75 |
**Size:** 12,606 prompts
|
|
|
|
| 76 |
Each row in the CSV corresponds to a single prompt instance and includes the following fields:
|
| 77 |
|
| 78 |
- `type`: Prompt category — one of `A`, `B`, `C`, or `D`, based on number of objects and complexity.
|
|
|
|
| 72 |
## Dataset Structure
|
| 73 |
|
| 74 |
### CSV Format (`prompt2scene_prompts_final.csv`)
|
| 75 |
+
|
| 76 |
**Size:** 12,606 prompts
|
| 77 |
+
|
| 78 |
Each row in the CSV corresponds to a single prompt instance and includes the following fields:
|
| 79 |
|
| 80 |
- `type`: Prompt category — one of `A`, `B`, `C`, or `D`, based on number of objects and complexity.
|