Guardado en:
Detalles Bibliográficos
Autores principales: Ghatkesar, Aarti, Venkatesh, Ganesh
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2505.05626
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909671299743744
author Ghatkesar, Aarti
Venkatesh, Ganesh
author_facet Ghatkesar, Aarti
Venkatesh, Ganesh
contents Achieving deep alignment between vision and language remains a central challenge for Multimodal Large Language Models (MLLMs). These models often fail to fully leverage visual input, defaulting to strong language priors. Our approach first provides insights into how MLLMs internally build visual understanding of image regions and then introduces techniques to amplify this capability. Specifically, we explore techniques designed both to deepen the model's understanding of visual content and to ensure that these visual insights actively guide language generation. We demonstrate the superior multimodal understanding of our resultant model through a detailed upstream analysis quantifying its ability to predict visually-dependent tokens as well as 10 pt boost on visually challenging tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2505_05626
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Perceiving Beyond Language Priors: Enhancing Visual Comprehension and Attention in Multimodal Models
Ghatkesar, Aarti
Venkatesh, Ganesh
Computer Vision and Pattern Recognition
Artificial Intelligence
Achieving deep alignment between vision and language remains a central challenge for Multimodal Large Language Models (MLLMs). These models often fail to fully leverage visual input, defaulting to strong language priors. Our approach first provides insights into how MLLMs internally build visual understanding of image regions and then introduces techniques to amplify this capability. Specifically, we explore techniques designed both to deepen the model's understanding of visual content and to ensure that these visual insights actively guide language generation. We demonstrate the superior multimodal understanding of our resultant model through a detailed upstream analysis quantifying its ability to predict visually-dependent tokens as well as 10 pt boost on visually challenging tasks.
title Perceiving Beyond Language Priors: Enhancing Visual Comprehension and Attention in Multimodal Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2505.05626