MLLMs Need 3D-Aware Representation Supervision for Scene Understanding

Xiaohu Huang, Jingjing Wu, Qunyi Xie, Kai Han

Keywords: Scene Understanding, Multimodal Large Language Models

TL;DR: Improving 3D representation learning for multimodal large language models.

Abstract: Recent advances in scene understanding have leveraged multimodal large language models (MLLMs) for 3D reasoning by capitalizing on their strong 2D pretraining. However, the lack of explicit 3D data during MLLM pretraining limits 3D representation capability. In this paper, we investigate the 3D-awareness of MLLMs by evaluating multi-view correspondence and reveal a strong positive correlation between the quality of 3D-aware representation and downstream task performance. Motivated by this, we propose 3DRS, a framework that enhances MLLM 3D representation learning by introducing supervision from pretrained 3D foundation models. Our approach aligns MLLM visual features with rich 3D knowledge distilled from 3D models, effectively improving scene understanding. Extensive experiments across multiple benchmarks and MLLMs—including visual grounding, captioning, and question answering—demonstrate consistent performance gains. Code will be released to facilitate future research.

Primary Area: Deep learning (e.g., architectures, generative models, optimization for deep networks, foundation models, LLMs)

Submission Number: 1735