CVMar 22, 2025

4D-Bench: Benchmarking Multi-modal Large Language Models for 4D Object Understanding

arXiv:2503.17827v14 citationsh-index: 24Has Code
Originality Synthesis-oriented
AI Analysis

This work addresses a gap in benchmarking for AI researchers and developers by providing a new benchmark for 4D object understanding, though it is incremental as it extends existing 2D/video benchmarks to a new domain.

The authors tackled the lack of standardized benchmarks for evaluating multimodal large language models (MLLMs) in understanding 4D objects (3D objects with temporal evolution) by introducing 4D-Bench, which includes tasks like 4D object QA and captioning, and found that MLLMs perform poorly, with GPT-4o achieving only 63% accuracy compared to a human baseline of 91%.

Multimodal Large Language Models (MLLMs) have demonstrated impressive 2D image/video understanding capabilities. However, there are no publicly standardized benchmarks to assess the abilities of MLLMs in understanding the 4D objects (3D objects with temporal evolution over time). In this paper, we introduce 4D-Bench, the first benchmark to evaluate the capabilities of MLLMs in 4D object understanding, featuring tasks in 4D object Question Answering (4D object QA) and 4D object captioning. 4D-Bench provides 4D objects with diverse categories, high-quality annotations, and tasks necessitating multi-view spatial-temporal understanding, different from existing 2D image/video-based benchmarks. With 4D-Bench, we evaluate a wide range of open-source and closed-source MLLMs. The results from the 4D object captioning experiment indicate that MLLMs generally exhibit weaker temporal understanding compared to their appearance understanding, notably, while open-source models approach closed-source performance in appearance understanding, they show larger performance gaps in temporal understanding. 4D object QA yields surprising findings: even with simple single-object videos, MLLMs perform poorly, with state-of-the-art GPT-4o achieving only 63\% accuracy compared to the human baseline of 91\%. These findings highlight a substantial gap in 4D object understanding and the need for further advancements in MLLMs.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes