(Almost) Free Modality Stitching of Foundation Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Singh, Jaisidh, Misra, Diganta, Knyazev, Boris, Orvieto, Antonio
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912487619690496
author Singh, Jaisidh
Misra, Diganta
Knyazev, Boris
Orvieto, Antonio
author_facet Singh, Jaisidh
Misra, Diganta
Knyazev, Boris
Orvieto, Antonio
contents Foundation multi-modal models are often designed by stitching of multiple existing pretrained uni-modal models: for example, an image classifier with an text model. This stitching process is performed by training a connector module that aims to align the representation spaces of these uni-modal models towards a multi-modal objective. However, given the complexity of training such connectors on large scale web-based datasets coupled with the ever-increasing number of available pretrained uni-modal models, the task of uni-modal models selection and subsequent connector module training becomes computationally demanding. To address this under-studied critical problem, we propose Hypernetwork Model Alignment (Hyma), a novel all-in-one solution for optimal uni-modal model selection and connector training by leveraging hypernetworks. Specifically, our framework utilizes the parameter prediction capability of a hypernetwork to obtain jointly trained connector modules for $N \times M$ combinations of uni-modal models. In our experiments, Hyma reduces the cost of searching for the best performing uni-modal model pair by $10\times$, while matching the ranking and trained connector performance obtained via grid search across a suite of diverse multi-modal benchmarks.
format Preprint
id arxiv_https___arxiv_org_abs_2507_10015
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle (Almost) Free Modality Stitching of Foundation Models
Singh, Jaisidh
Misra, Diganta
Knyazev, Boris
Orvieto, Antonio
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Foundation multi-modal models are often designed by stitching of multiple existing pretrained uni-modal models: for example, an image classifier with an text model. This stitching process is performed by training a connector module that aims to align the representation spaces of these uni-modal models towards a multi-modal objective. However, given the complexity of training such connectors on large scale web-based datasets coupled with the ever-increasing number of available pretrained uni-modal models, the task of uni-modal models selection and subsequent connector module training becomes computationally demanding. To address this under-studied critical problem, we propose Hypernetwork Model Alignment (Hyma), a novel all-in-one solution for optimal uni-modal model selection and connector training by leveraging hypernetworks. Specifically, our framework utilizes the parameter prediction capability of a hypernetwork to obtain jointly trained connector modules for $N \times M$ combinations of uni-modal models. In our experiments, Hyma reduces the cost of searching for the best performing uni-modal model pair by $10\times$, while matching the ranking and trained connector performance obtained via grid search across a suite of diverse multi-modal benchmarks.
title (Almost) Free Modality Stitching of Foundation Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2507.10015