ComCLIP: Training-Free Compositional Image and Text Matching

Contrastive Language-Image Pretraining (CLIP) has demonstrated great zero-shot performance for matching images and text. However, it is still challenging to adapt vision-lanaguage pretrained models like CLIP to compositional image and text matching -- a more challenging image and text matching task requiring the model understanding of compositional word concepts and visual components. Towards better compositional generalization in zero-shot image and text matching, in this paper, we study the problem from a causal perspective: the erroneous semantics of individual entities are essentially confounders that cause the matching failure. Therefore, we propose a novel textbftextittraining-free compositional CLIP model (ComCLIP). ComCLIP disentangles input images into subjects, objects, and action sub-images and composes CLIP's vision encoder and text encoder to perform evolving matching over compositional text embedding and sub-image embeddings. In this way, ComCLIP can mitigate spurious correlations introduced by the pretrained CLIP models and dynamically evaluate the importance of each component. Experiments on four compositional image-text matching datasets: SVO, ComVG, Winoground, and VL-checklist, and two general image-text retrieval datasets: Flick30K, and MSCOCO demonstrate the effectiveness of our plug-and-play method, which boosts the textbftextitzero-shot inference ability of CLIP, SLIP, and BLIP2 even without further training or fine-tuning. Our codes can be found at https://github.com/eric-ai-lab/ComCLIP.