They're All Doctors: Synthesizing Diverse Counterfactuals to Mitigate Associative Bias
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929388047564800 |
|---|---|
| author | Magid, Salma Abdel Wang, Jui-Hsien Kafle, Kushal Pfister, Hanspeter |
| author_facet | Magid, Salma Abdel Wang, Jui-Hsien Kafle, Kushal Pfister, Hanspeter |
| contents | Vision Language Models (VLMs) such as CLIP are powerful models; however they can exhibit unwanted biases, making them less safe when deployed directly in applications such as text-to-image, text-to-video retrievals, reverse search, or classification tasks. In this work, we propose a novel framework to generate synthetic counterfactual images to create a diverse and balanced dataset that can be used to fine-tune CLIP. Given a set of diverse synthetic base images from text-to-image models, we leverage off-the-shelf segmentation and inpainting models to place humans with diverse visual appearances in context. We show that CLIP trained on such datasets learns to disentangle the human appearance from the context of an image, i.e., what makes a doctor is not correlated to the person's visual appearance, like skin color or body type, but to the context, such as background, the attire they are wearing, or the objects they are holding. We demonstrate that our fine-tuned CLIP model, $CF_α$, improves key fairness metrics such as MaxSkew, MinSkew, and NDKL by 40-66\% for image retrieval tasks, while still achieving similar levels of performance in downstream tasks. We show that, by design, our model retains maximal compatibility with the original CLIP models, and can be easily controlled to support different accuracy versus fairness trade-offs in a plug-n-play fashion. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_11331 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | They're All Doctors: Synthesizing Diverse Counterfactuals to Mitigate Associative Bias Magid, Salma Abdel Wang, Jui-Hsien Kafle, Kushal Pfister, Hanspeter Computer Vision and Pattern Recognition Information Retrieval Machine Learning Vision Language Models (VLMs) such as CLIP are powerful models; however they can exhibit unwanted biases, making them less safe when deployed directly in applications such as text-to-image, text-to-video retrievals, reverse search, or classification tasks. In this work, we propose a novel framework to generate synthetic counterfactual images to create a diverse and balanced dataset that can be used to fine-tune CLIP. Given a set of diverse synthetic base images from text-to-image models, we leverage off-the-shelf segmentation and inpainting models to place humans with diverse visual appearances in context. We show that CLIP trained on such datasets learns to disentangle the human appearance from the context of an image, i.e., what makes a doctor is not correlated to the person's visual appearance, like skin color or body type, but to the context, such as background, the attire they are wearing, or the objects they are holding. We demonstrate that our fine-tuned CLIP model, $CF_α$, improves key fairness metrics such as MaxSkew, MinSkew, and NDKL by 40-66\% for image retrieval tasks, while still achieving similar levels of performance in downstream tasks. We show that, by design, our model retains maximal compatibility with the original CLIP models, and can be easily controlled to support different accuracy versus fairness trade-offs in a plug-n-play fashion. |
| title | They're All Doctors: Synthesizing Diverse Counterfactuals to Mitigate Associative Bias |
| topic | Computer Vision and Pattern Recognition Information Retrieval Machine Learning |
| url | https://arxiv.org/abs/2406.11331 |