SAVVY: Spatial Awareness via Audio-Visual LLMs through Seeing and Hearing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Mingfei, Cui, Zijun, Liu, Xiulong, Xiang, Jinlin, Zheng, Caleb, Li, Jingyuan, Shlizerman, Eli
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918046553079808
author Chen, Mingfei
Cui, Zijun
Liu, Xiulong
Xiang, Jinlin
Zheng, Caleb
Li, Jingyuan
Shlizerman, Eli
author_facet Chen, Mingfei
Cui, Zijun
Liu, Xiulong
Xiang, Jinlin
Zheng, Caleb
Li, Jingyuan
Shlizerman, Eli
contents 3D spatial reasoning in dynamic, audio-visual environments is a cornerstone of human cognition yet remains largely unexplored by existing Audio-Visual Large Language Models (AV-LLMs) and benchmarks, which predominantly focus on static or 2D scenes. We introduce SAVVY-Bench, the first benchmark for 3D spatial reasoning in dynamic scenes with synchronized spatial audio. SAVVY-Bench is comprised of thousands of relationships involving static and moving objects, and requires fine-grained temporal grounding, consistent 3D localization, and multi-modal annotation. To tackle this challenge, we propose SAVVY, a novel training-free reasoning pipeline that consists of two stages: (i) Egocentric Spatial Tracks Estimation, which leverages AV-LLMs as well as other audio-visual methods to track the trajectories of key objects related to the query using both visual and spatial audio cues, and (ii) Dynamic Global Map Construction, which aggregates multi-modal queried object trajectories and converts them into a unified global dynamic map. Using the constructed map, a final QA answer is obtained through a coordinate transformation that aligns the global map with the queried viewpoint. Empirical evaluation demonstrates that SAVVY substantially enhances performance of state-of-the-art AV-LLMs, setting a new standard and stage for approaching dynamic 3D spatial reasoning in AV-LLMs.
format Preprint
id arxiv_https___arxiv_org_abs_2506_05414
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SAVVY: Spatial Awareness via Audio-Visual LLMs through Seeing and Hearing
Chen, Mingfei
Cui, Zijun
Liu, Xiulong
Xiang, Jinlin
Zheng, Caleb
Li, Jingyuan
Shlizerman, Eli
Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Multimedia
Sound
Audio and Speech Processing
3D spatial reasoning in dynamic, audio-visual environments is a cornerstone of human cognition yet remains largely unexplored by existing Audio-Visual Large Language Models (AV-LLMs) and benchmarks, which predominantly focus on static or 2D scenes. We introduce SAVVY-Bench, the first benchmark for 3D spatial reasoning in dynamic scenes with synchronized spatial audio. SAVVY-Bench is comprised of thousands of relationships involving static and moving objects, and requires fine-grained temporal grounding, consistent 3D localization, and multi-modal annotation. To tackle this challenge, we propose SAVVY, a novel training-free reasoning pipeline that consists of two stages: (i) Egocentric Spatial Tracks Estimation, which leverages AV-LLMs as well as other audio-visual methods to track the trajectories of key objects related to the query using both visual and spatial audio cues, and (ii) Dynamic Global Map Construction, which aggregates multi-modal queried object trajectories and converts them into a unified global dynamic map. Using the constructed map, a final QA answer is obtained through a coordinate transformation that aligns the global map with the queried viewpoint. Empirical evaluation demonstrates that SAVVY substantially enhances performance of state-of-the-art AV-LLMs, setting a new standard and stage for approaching dynamic 3D spatial reasoning in AV-LLMs.
title SAVVY: Spatial Awareness via Audio-Visual LLMs through Seeing and Hearing
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Machine Learning
Multimedia
Sound
Audio and Speech Processing
url https://arxiv.org/abs/2506.05414