Finetuning CLIP to Reason about Pairwise Differences

• URL: http://arxiv.org/abs/2409.09721v1
• Date: Sun, 15 Sep 2024 13:02:14 GMT
• Title: Finetuning CLIP to Reason about Pairwise Differences
• Authors: Dylan Sam, Devin Willmott, Joao D. Semedo, J. Zico Kolter,
• Abstract summary: We propose an approach to train vision-language models such as CLIP in a contrastive manner to reason about differences in embedding space. We first demonstrate that our approach yields significantly improved capabilities in ranking images by a certain attribute. We also illustrate that the resulting embeddings obey a larger degree of geometric properties in embedding space.
• Score: 52.028073305958074
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision-language models (VLMs) such as CLIP are trained via contrastive learning between text and image pairs, resulting in aligned image and text embeddings that are useful for many downstream tasks. A notable drawback of CLIP, however, is that the resulting embedding space seems to lack some of the structure of their purely text-based alternatives. For instance, while text embeddings have been long noted to satisfy \emph{analogies} in embedding space using vector arithmetic, CLIP has no such property. In this paper, we propose an approach to natively train CLIP in a contrastive manner to reason about differences in embedding space. We finetune CLIP so that the differences in image embedding space correspond to \emph{text descriptions of the image differences}, which we synthetically generate with large language models on image-caption paired datasets. We first demonstrate that our approach yields significantly improved capabilities in ranking images by a certain attribute (e.g., elephants are larger than cats), which is useful in retrieval or constructing attribute-based classifiers, and improved zeroshot classification performance on many downstream image classification tasks. In addition, our approach enables a new mechanism for inference that we refer to as comparative prompting, where we leverage prior knowledge of text descriptions of differences between classes of interest, achieving even larger performance gains in classification. Finally, we illustrate that the resulting embeddings obey a larger degree of geometric properties in embedding space, such as in text-to-image generation.

Related papers

• TripletCLIP: Improving Compositional Reasoning of CLIP via Synthetic Vision-Language Negatives [65.82577305915643]
  Contrastive Language-Image Pretraining (CLIP) models maximize the mutual information between text and visual modalities to learn representations. We show that generating hard'' negative captions via in-context learning and corresponding negative images with text-to-image generators offers a solution. We demonstrate that our method, named TripletCLIP, enhances the compositional capabilities of CLIP, resulting in an absolute improvement of over 9% on the SugarCrepe benchmark.
  arXiv  Detail & Related papers  (2024-11-04T19:24:59Z)
• Selective Vision-Language Subspace Projection for Few-shot CLIP [55.361337202198925]
  We introduce a method called Selective Vision-Language Subspace Projection (SSP) SSP incorporates local image features and utilizes them as a bridge to enhance the alignment between image-text pairs. Our approach entails only training-free matrix calculations and can be seamlessly integrated into advanced CLIP-based few-shot learning frameworks.
  arXiv  Detail & Related papers  (2024-07-24T03:45:35Z)
• STAIR: Learning Sparse Text and Image Representation in Grounded Tokens [84.14528645941128]
  We show that it is possible to build a sparse semantic representation that is as powerful as, or even better than, dense presentations. We extend the CLIP model and build a sparse text and image representation (STAIR), where the image and text are mapped to a sparse token space. It significantly outperforms a CLIP model with +$4.9%$ and +$4.3%$ absolute Recall@1 improvement.
  arXiv  Detail & Related papers  (2023-01-30T17:21:30Z)
• CyCLIP: Cyclic Contrastive Language-Image Pretraining [34.588147979731374]
  Recent advances in contrastive representation learning over paired image-text data have led to models such as CLIP that achieve state-of-the-art performance for zero-shot classification and distributional robustness. We demonstrate that the image and text representations learned via a standard contrastive objective are not interchangeable and can lead to inconsistent downstream predictions. We propose CyCLIP, a framework for contrastive representation learning that explicitly optimize for the learned representations to be geometrically consistent in the image and text space.
  arXiv  Detail & Related papers  (2022-05-28T15:31:17Z)
• No Token Left Behind: Explainability-Aided Image Classification and Generation [79.4957965474334]
  We present a novel explainability-based approach, which adds a loss term to ensure that CLIP focuses on all relevant semantic parts of the input. Our method yields an improvement in the recognition rate, without additional training or fine-tuning.
  arXiv  Detail & Related papers  (2022-04-11T07:16:39Z)
• VT-CLIP: Enhancing Vision-Language Models with Visual-guided Texts [2.0434814235659555]
  Contrastive Language-Image Pre-training (CLIP) has drawn increasing attention recently for its transferable visual representation learning. We propose to enhance CLIP via Visual-guided Texts, named VT-CLIP. In few-shot settings, we evaluate our VT-CLIP on 11 well-known classification datasets to demonstrate its effectiveness.
  arXiv  Detail & Related papers  (2021-12-04T18:34:24Z)
• CRIS: CLIP-Driven Referring Image Segmentation [71.56466057776086]
  We propose an end-to-end CLIP-Driven Referring Image framework (CRIS) CRIS resorts to vision-language decoding and contrastive learning for achieving the text-to-pixel alignment. Our proposed framework significantly outperforms the state-of-the-art performance without any post-processing.
  arXiv  Detail & Related papers  (2021-11-30T07:29:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.