Do Vision-Language Models Really Understand Visual Language?

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Hou, Yifan, Giledereli, Buse, Tu, Yilei, Sachan, Mrinmaya
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913856444432384
author Hou, Yifan
Giledereli, Buse
Tu, Yilei
Sachan, Mrinmaya
author_facet Hou, Yifan
Giledereli, Buse
Tu, Yilei
Sachan, Mrinmaya
contents Visual language is a system of communication that conveys information through symbols, shapes, and spatial arrangements. Diagrams are a typical example of a visual language depicting complex concepts and their relationships in the form of an image. The symbolic nature of diagrams presents significant challenges for building models capable of understanding them. Recent studies suggest that Large Vision-Language Models (LVLMs) can even tackle complex reasoning tasks involving diagrams. In this paper, we investigate this phenomenon by developing a comprehensive test suite to evaluate the diagram comprehension capability of LVLMs. Our test suite uses a variety of questions focused on concept entities and their relationships over a set of synthetic as well as real diagrams across domains to evaluate the recognition and reasoning abilities of models. Our evaluation of LVLMs shows that while they can accurately identify and reason about entities, their ability to understand relationships is notably limited. Further testing reveals that the decent performance on diagram understanding largely stems from leveraging their background knowledge as shortcuts to identify and reason about the relational information. Thus, we conclude that LVLMs have a limited capability for genuine diagram understanding, and their impressive performance in diagram reasoning is an illusion emanating from other confounding factors, such as the background knowledge in the models.
format Preprint
id arxiv_https___arxiv_org_abs_2410_00193
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Do Vision-Language Models Really Understand Visual Language?
Hou, Yifan
Giledereli, Buse
Tu, Yilei
Sachan, Mrinmaya
Computation and Language
Computer Vision and Pattern Recognition
Visual language is a system of communication that conveys information through symbols, shapes, and spatial arrangements. Diagrams are a typical example of a visual language depicting complex concepts and their relationships in the form of an image. The symbolic nature of diagrams presents significant challenges for building models capable of understanding them. Recent studies suggest that Large Vision-Language Models (LVLMs) can even tackle complex reasoning tasks involving diagrams. In this paper, we investigate this phenomenon by developing a comprehensive test suite to evaluate the diagram comprehension capability of LVLMs. Our test suite uses a variety of questions focused on concept entities and their relationships over a set of synthetic as well as real diagrams across domains to evaluate the recognition and reasoning abilities of models. Our evaluation of LVLMs shows that while they can accurately identify and reason about entities, their ability to understand relationships is notably limited. Further testing reveals that the decent performance on diagram understanding largely stems from leveraging their background knowledge as shortcuts to identify and reason about the relational information. Thus, we conclude that LVLMs have a limited capability for genuine diagram understanding, and their impressive performance in diagram reasoning is an illusion emanating from other confounding factors, such as the background knowledge in the models.
title Do Vision-Language Models Really Understand Visual Language?
topic Computation and Language
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.00193