Vision-Language Models Performing Zero-Shot Tasks Exhibit Gender-based Disparities
We explore the extent to which zero-shot vision-language models exhibit gender bias for different vision tasks. Vision models traditionally required task-specific labels for representing concepts, as well as finetuning; zero-shot models like CLIP instead perform tasks with an open-vocabulary, meaning they do not need a fixed set of labels, by using text embeddings to represent concepts. With these capabilities in mind, we ask: Do vision-language models exhibit gender bias when performing zero-shot image classification, object detection and semantic segmentation? We evaluate different vision-language models with multiple datasets across a set of concepts and find (i) all models evaluated show distinct performance differences based on the perceived gender of the person co-occurring with a given concept in the image and that aggregating analyses over all concepts can mask these concerns; (ii) model calibration (i.e. the relationship between accuracy and confidence) also differs distinctly by perceived gender, even when evaluating on similar representations of concepts; and (iii) these observed disparities align with existing gender biases in word embeddings from language models. These findings suggest that, while language greatly expands the capability of vision tasks, it can also contribute to social biases in zero-shot vision settings. Furthermore, biases can further propagate when foundational models like CLIP are used by other models to enable zero-shot capabilities.
Code (0)
등록된 구현이 없습니다.
Tasks
image-classificationImage Classificationobject-detectionObject DetectionSemantic SegmentationWord EmbeddingsZero-Shot Image ClassificationMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Filtering, Distillation, and Hard Negatives for Vision-Language Pre-Training
Vision-language models trained with contrastive learning on large-scale noisy data are becoming increasingly popular for zero-shot recognition problems. In this paper we improve the following three aspects of the contras…
Contrastive LearningText SpottingZero-Shot LearningOn the Zero-shot Adversarial Robustness of Vision-Language Models: A Truly Zero-shot and Training-free Approach
Pre-trained Vision-Language Models (VLMs) like CLIP, have demonstrated strong zero-shot generalization capabilities. Despite their effectiveness on various downstream tasks, they remain vulnerable to adversarial samp…
Adversarial RobustnessZero-shot GeneralizationLet's Do a Thought Experiment: Using Counterfactuals to Improve Moral Reasoning
Language models still struggle on moral reasoning, despite their impressive performance in many other tasks. In particular, the Moral Scenarios task in MMLU (Multi-task Language Understanding) is among the worst performi…
counterfactualMathMMLUMoral Scenarios+1ImageBind: One Embedding Space To Bind Them All
We present ImageBind, an approach to learn a joint embedding across six different modalities - images, text, audio, depth, thermal, and IMU data. We show that all combinations of paired data are not necessary to train su…
AllCross-Modal RetrievalMultimodal Deep LearningRetrieval+10Modeling Caption Diversity in Contrastive Vision-Language Pretraining
There are a thousand ways to caption an image. Contrastive Language Pretraining (CLIP) on the other hand, works by mapping an image and its caption to a single vector -- limiting how well CLIP-like models can represent t…
Diversityzero-shot-classificationZero-Shot Learning