UNICBench: UNIfied Counting Benchmark for MLLM
Rong, Chenggang, Han, Tao, Zhao, Zhiyuan, Fan, Yaowu, Wan, Jia, Guo, Song, Yuan, Yuan, Gao, Junyu
原文摘要
Counting is a core capability for multimodal large language models (MLLMs), yet there is no unified counting dataset to rigorously evaluate this ability across image, text, and audio. We present UNICBench, a unified multimodal, multi-level counting benchmark and evaluation toolkit with accurate ground truth, deterministic numeric parsing, and stratified reporting. The corpus comprises 5,300 images (5,508 QA), 872 documents (5,888 QA), and 2,069 audio clips (2,905 QA), annotated with a three-level capability taxonomy and difficulty tags. Under a standardized protocol with fixed splits/prompts/seeds and modality-specific matching rules, we evaluate 45 state-of-the-art MLLMs across modalities. Results show strong performance on some basic counting tasks but significant gaps on reasoning and the hardest partitions, highlighting long-tail errors and substantial headroom for improving general counting. UNICBench offers a rigorous and comparable basis for measurement and a public toolkit to accelerate progress.
cs.CV · cs.CL
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition · 2026 · 正式发表