Beyond Sight: Towards Cognitive Alignment in LVLM via Enriched Visual Knowledge

TL;DR

This paper investigates the cognitive misalignment in large vision-language models caused by visual encoder limitations and proposes an entity-enhanced method to improve alignment and landmark recognition performance.

Contribution

It introduces a multi-granularity supervision approach, Entity-Enhanced Cognitive Alignment (EECA), to generate visually enriched tokens that better align with the LLM's cognitive framework.

Findings

VE-Unknown data hampers LVLM understanding

Rich visual features in VE-Known data improve alignment

EECA significantly boosts landmark recognition accuracy

Abstract

Does seeing always mean knowing? Large Vision-Language Models (LVLMs) integrate separately pre-trained vision and language components, often using CLIP-ViT as vision backbone. However, these models frequently encounter a core issue of "cognitive misalignment" between the vision encoder (VE) and the large language model (LLM). Specifically, the VE's representation of visual information may not fully align with LLM's cognitive framework, leading to a mismatch where visual features exceed the language model's interpretive range. To address this, we investigate how variations in VE representations influence LVLM comprehension, especially when the LLM faces VE-Unknown data-images whose ambiguous visual representations challenge the VE's interpretive precision. Accordingly, we construct a multi-granularity landmark dataset and systematically examine the impact of VE-Known and VE-Unknown data on interpretive abilities. Our results show that VE-Unknown data limits LVLM's capacity for accurate understanding, while VE-Known data, rich in distinctive features, helps reduce cognitive misalignment. Building on these insights, we propose Entity-Enhanced Cognitive Alignment (EECA), a method that employs multi-granularity supervision to generate visually enriched, well-aligned tokens that not only integrate within the LLM's embedding space but also align with the LLM's cognitive framework. This alignment markedly enhances LVLM performance in landmark recognition. Our findings underscore the challenges posed by VE-Unknown data and highlight the essential role of cognitive alignment in advancing multimodal systems.