Datasets:
Update README.md
Browse files
README.md
CHANGED
|
@@ -73,6 +73,8 @@ We asked humans to evaluate AI-generated images in style, coherence and prompt a
|
|
| 73 |
|
| 74 |
If you want to replicate the annotation setup, the steps are outlined at the [bottom](#replicating-the-annotation-setup).
|
| 75 |
|
|
|
|
|
|
|
| 76 |
# Word Scores
|
| 77 |
Users identified words from the prompts that were NOT accurately depicted in the generated images. Higher word scores indicate poorer representation in the image. Participants also had the option to select "[No_mistakes]" for prompts where all elements were accurately depicted.
|
| 78 |
|
|
|
|
| 73 |
|
| 74 |
If you want to replicate the annotation setup, the steps are outlined at the [bottom](#replicating-the-annotation-setup).
|
| 75 |
|
| 76 |
+
This dataset and the annotation process is described in further detail in our blog post [Beyond Image Preferences](https://huggingface.co/blog/RapidataAI/beyond-image-preferences).
|
| 77 |
+
|
| 78 |
# Word Scores
|
| 79 |
Users identified words from the prompts that were NOT accurately depicted in the generated images. Higher word scores indicate poorer representation in the image. Participants also had the option to select "[No_mistakes]" for prompts where all elements were accurately depicted.
|
| 80 |
|