MM-R$^3$: On (In-)Consistency of Vision-Language Models (VLMs)

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chou, Shih-Han, Chandhok, Shivam, Little, James J., Sigal, Leonid
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918073154404352
author Chou, Shih-Han
Chandhok, Shivam
Little, James J.
Sigal, Leonid
author_facet Chou, Shih-Han
Chandhok, Shivam
Little, James J.
Sigal, Leonid
contents With the advent of LLMs and variants, a flurry of research has emerged, analyzing the performance of such models across an array of tasks. While most studies focus on evaluating the capabilities of state-of-the-art (SoTA) Vision Language Models (VLMs) through task accuracy (e.g., visual question answering, grounding), our work explores the related but complementary aspect of consistency - the ability of a VLM to produce semantically similar or identical responses to semantically similar queries. We note that consistency is a fundamental prerequisite (necessary but not sufficient condition) for robustness and trust in VLMs. Armed with this perspective, we propose the MM-R3 benchmark, which allows us to analyze performance, in terms of consistency and accuracy, of SoTA VLMs on three tasks: Question Rephrasing, Image Restyling, and Context Reasoning. Our analysis reveals that consistency does not always align with accuracy, indicating that models with higher accuracy are not necessarily more consistent, and vice versa. Furthermore, we propose a simple yet effective mitigation strategy in the form of an adapter module trained to minimize inconsistency across prompts. With our proposed strategy, we are able to achieve absolute improvements of 5.7% and 12.5%, on average on widely used VLMs such as BLIP-2 and LLaVa 1.5M in terms of consistency over their existing counterparts.
format Preprint
id arxiv_https___arxiv_org_abs_2410_04778
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle MM-R$^3$: On (In-)Consistency of Vision-Language Models (VLMs)
Chou, Shih-Han
Chandhok, Shivam
Little, James J.
Sigal, Leonid
Computer Vision and Pattern Recognition
With the advent of LLMs and variants, a flurry of research has emerged, analyzing the performance of such models across an array of tasks. While most studies focus on evaluating the capabilities of state-of-the-art (SoTA) Vision Language Models (VLMs) through task accuracy (e.g., visual question answering, grounding), our work explores the related but complementary aspect of consistency - the ability of a VLM to produce semantically similar or identical responses to semantically similar queries. We note that consistency is a fundamental prerequisite (necessary but not sufficient condition) for robustness and trust in VLMs. Armed with this perspective, we propose the MM-R3 benchmark, which allows us to analyze performance, in terms of consistency and accuracy, of SoTA VLMs on three tasks: Question Rephrasing, Image Restyling, and Context Reasoning. Our analysis reveals that consistency does not always align with accuracy, indicating that models with higher accuracy are not necessarily more consistent, and vice versa. Furthermore, we propose a simple yet effective mitigation strategy in the form of an adapter module trained to minimize inconsistency across prompts. With our proposed strategy, we are able to achieve absolute improvements of 5.7% and 12.5%, on average on widely used VLMs such as BLIP-2 and LLaVa 1.5M in terms of consistency over their existing counterparts.
title MM-R$^3$: On (In-)Consistency of Vision-Language Models (VLMs)
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.04778