Add dataset card and link to paper
#2
by nielsr HF Staff - opened
README.md
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
task_categories:
|
| 3 |
+
- image-classification
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# ImageNet-100-Text
|
| 7 |
+
|
| 8 |
+
This dataset is used to evaluate the robustness of CLIP and vision-language models (VLMs) against typographic attacks (where irrelevant text overlaid on images can mislead visual representations).
|
| 9 |
+
|
| 10 |
+
It is introduced and evaluated in the paper [Towards Robustness against Typographic Attack with Training-free Concept Localization](https://huggingface.co/papers/2607.02494).
|
| 11 |
+
|
| 12 |
+
## Links
|
| 13 |
+
|
| 14 |
+
- **Paper**: [Towards Robustness against Typographic Attack with Training-free Concept Localization](https://huggingface.co/papers/2607.02494)
|
| 15 |
+
- **Code**: [GitHub - Liu-524/SamplingTAR](https://github.com/Liu-524/SamplingTAR)
|