HAMMER: Harnessing MLLM via Cross-Modal Integration for Intention-Driven 3D Affordance Grounding

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yao, Lei, Chen, Yong, Su, Yuejiao, Wang, Yi, Liu, Moyun, Chau, Lap-Pui
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912938735960064
author Yao, Lei
Chen, Yong
Su, Yuejiao
Wang, Yi
Liu, Moyun
Chau, Lap-Pui
author_facet Yao, Lei
Chen, Yong
Su, Yuejiao
Wang, Yi
Liu, Moyun
Chau, Lap-Pui
contents Humans commonly identify 3D object affordance through observed interactions in images or videos, and once formed, such knowledge can be generically generalized to novel objects. Inspired by this principle, we advocate for a novel framework that leverages emerging multimodal large language models (MLLMs) for interaction intention-driven 3D affordance grounding, namely HAMMER. Instead of generating explicit object attribute descriptions or relying on off-the-shelf 2D segmenters, we alternatively aggregate the interaction intention depicted in the image into a contact-aware embedding and guide the model to infer textual affordance labels, ensuring it thoroughly excavates object semantics and contextual cues. We further devise a hierarchical cross-modal integration mechanism to fully exploit the complementary information from the MLLM for 3D representation refinement and introduce a multi-granular geometry lifting module that infuses spatial characteristics into the extracted intention embedding, thus facilitating accurate 3D affordance localization. Extensive experiments on public datasets and our newly constructed corrupted benchmark demonstrate the superiority and robustness of HAMMER compared to existing approaches. All code and weights are publicly available.
format Preprint
id arxiv_https___arxiv_org_abs_2603_02329
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle HAMMER: Harnessing MLLM via Cross-Modal Integration for Intention-Driven 3D Affordance Grounding
Yao, Lei
Chen, Yong
Su, Yuejiao
Wang, Yi
Liu, Moyun
Chau, Lap-Pui
Computer Vision and Pattern Recognition
Humans commonly identify 3D object affordance through observed interactions in images or videos, and once formed, such knowledge can be generically generalized to novel objects. Inspired by this principle, we advocate for a novel framework that leverages emerging multimodal large language models (MLLMs) for interaction intention-driven 3D affordance grounding, namely HAMMER. Instead of generating explicit object attribute descriptions or relying on off-the-shelf 2D segmenters, we alternatively aggregate the interaction intention depicted in the image into a contact-aware embedding and guide the model to infer textual affordance labels, ensuring it thoroughly excavates object semantics and contextual cues. We further devise a hierarchical cross-modal integration mechanism to fully exploit the complementary information from the MLLM for 3D representation refinement and introduce a multi-granular geometry lifting module that infuses spatial characteristics into the extracted intention embedding, thus facilitating accurate 3D affordance localization. Extensive experiments on public datasets and our newly constructed corrupted benchmark demonstrate the superiority and robustness of HAMMER compared to existing approaches. All code and weights are publicly available.
title HAMMER: Harnessing MLLM via Cross-Modal Integration for Intention-Driven 3D Affordance Grounding
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.02329