Can I read Cross the Gap: Exposing the Intra-modal Misalignment in CLIP via Modality Inversion on EtoBox?

Cross the Gap: Exposing the Intra-modal Misalignment in CLIP via Modality Inversion by Mistretta, Marco; Baldrati, Alberto; Agnolucci, Lorenzo; Bertini, Marco; Bagdanov, Andrew D. is a scholarly article available to read on EtoBox.

What is Cross the Gap: Exposing the Intra-modal Misalignment in CLIP via Modality Inversion about?

Pre-trained multi-modal Vision-Language Models like CLIP are widely used off-the-shelf for a variety of applications. In this paper, we show that the common practice of individually exploiting the text or image encoders of these powerful multi-modal models is highly suboptimal for intra-modal tasks like image-to-image retrieval. We argue that this is inherently due to the CLIP-style inter-modal contrastive loss that does not enforce any intra-modal constraints, leading to what we call intra-modal misalignment. To demonstrate this, we leverage two optimization-based modality inversion techniques that map representations from their input modality to the complementary one without any need for auxiliary data or additional trained adapters. We empirically show that, in the intra-modal tasks of image-to-image and text-to-text retrieval, approaching these tasks inter-modally significantly improves performance with respect to intra-modal baselines on more than fifteen datasets. Additionally, we demonstrate that approaching a native inter-modal task (e.g. zero-shot image classification) intra-modally decreases performance, further validating our findings. Finally, we show that incorporating

Author

Mistretta, Marco; Baldrati, Alberto; Agnolucci, Lorenzo; Bertini, Marco; Bagdanov, Andrew D.

Published

2025

Language

EN