Benchmarking Large Vision-Language Models on Fine-Grained Image Tasks: A Comprehensive Evaluation
This work addresses the problem of evaluating LVLMs on fine-grained tasks for researchers and developers, providing a comprehensive benchmark and guidance for future improvements, though it is incremental as it builds on existing evaluation studies.
The authors tackled the lack of evaluation for Large Vision-Language Models (LVLMs) on fine-grained image tasks by introducing FG-BMK, a benchmark with 1.01 million questions and 0.33 million images, and found key insights into training paradigms and performance limitations through experiments on twelve models.
Recent advancements in Large Vision-Language Models (LVLMs) have demonstrated remarkable multimodal perception capabilities, garnering significant attention. While numerous evaluation studies have emerged, assessing LVLMs both holistically and on specialized tasks, fine-grained image tasks-fundamental to computer vision-remain largely unexplored. To fill this gap, we introduce a comprehensive fine-grained evaluation benchmark, i.e., FG-BMK, comprising 1.01 million questions and 0.33 million images. Our evaluation systematically examines LVLMs from both human-oriented and machine-oriented perspectives, focusing on their semantic recognition and fine-grained feature representation capabilities. Through extensive experiments on twelve representative LVLMs/VLMs, we uncover key findings regarding the influence of training paradigms, modality alignment, perturbation susceptibility, and fine-grained category reasoning on task performance. This work provides critical insights into the limitations of current LVLMs and offers guidance for future data construction and model design in the development of more advanced LVLMs. Our code is open-source and available at https://github.com/SEU-VIPGroup/FG-BMK.