Benchmarking Large and Small MLLMs

TL;DR

This paper systematically benchmarks large and small multimodal language models, revealing that small models can match large ones in some tasks but struggle with complex reasoning, guiding future improvements.

Contribution

It provides a comprehensive evaluation of both large and small MLLMs across multiple capabilities and real-world scenarios, highlighting their strengths and limitations.

Findings

Small MLLMs perform comparably to large models in certain tasks.

Large models excel in complex reasoning and nuanced understanding.

Both model types exhibit common failure cases in specific domains.

Abstract

Large multimodal language models (MLLMs) such as GPT-4V and GPT-4o have achieved remarkable advancements in understanding and generating multimodal content, showcasing superior quality and capabilities across diverse tasks. However, their deployment faces significant challenges, including slow inference, high computational cost, and impracticality for on-device applications. In contrast, the emergence of small MLLMs, exemplified by the LLava-series models and Phi-3-Vision, offers promising alternatives with faster inference, reduced deployment costs, and the ability to handle domain-specific scenarios. Despite their growing presence, the capability boundaries between large and small MLLMs remain underexplored. In this work, we conduct a systematic and comprehensive evaluation to benchmark both small and large MLLMs, spanning general capabilities such as object recognition, temporal reasoning, and multimodal comprehension, as well as real-world applications in domains like industry and automotive. Our evaluation reveals that small MLLMs can achieve comparable performance to large models in specific scenarios but lag significantly in complex tasks requiring deeper reasoning or nuanced understanding. Furthermore, we identify common failure cases in both small and large MLLMs, highlighting domains where even state-of-the-art models struggle. We hope our findings will guide the research community in pushing the quality boundaries of MLLMs, advancing their usability and effectiveness across diverse applications.