PanoAffordanceNet: Towards Holistic Affordance Grounding in 360° Indoor Environments

2026-03-10

AI Summary

Existing affordance localization methods are constrained by perspective views and object-centric modeling, limiting their ability to support global perception for embodied agents in panoramic indoor environments. To address this gap, this work introduces the first scene-level panoramic affordance localization task and presents 360-AGD, the first high-quality panoramic affordance dataset. We propose PanoAffordanceNet, an end-to-end network featuring a Distortion-Aware Spectral Modulator (DASM) to correct equirectangular projection distortions and an Omnidirectional Spherical Densification Head (OSDH) to restore spherical topological continuity. The framework incorporates multi-level constraints spanning pixel-wise, distribution-level, and region-text contrastive learning. Experiments demonstrate that our approach significantly outperforms existing methods on 360-AGD, establishing a new baseline for scene-level perception in embodied intelligence.

Technology Category

• Intelligent Robots: Localization, Mapping, and Navigation
• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Systems and Infrastructure for Web, Mobile and WoT: Location- and context-aware Web and WoT applications and services
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Global perception is essential for embodied agents in 360° spaces, yet current affordance grounding remains largely object-centric and restricted to perspective views. To bridge this gap, we introduce a novel task: Holistic Affordance Grounding in 360° Indoor Environments. This task faces unique challenges, including severe geometric distortions from Equirectangular Projection (ERP), semantic dispersion, and cross-scale alignment difficulties. We propose PanoAffordanceNet, an end-to-end framework featuring a Distortion-Aware Spectral Modulator (DASM) for latitude-dependent calibration and an Omni-Spherical Densification Head (OSDH) to restore topological continuity from sparse activations. By integrating multi-level constraints comprising pixel-wise, distributional, and region-text contrastive objectives, our framework effectively suppresses semantic drift under low supervision. Furthermore, we construct 360-AGD, the first high-quality panoramic affordance grounding dataset. Extensive experiments demonstrate that PanoAffordanceNet significantly outperforms existing methods, establishing a solid baseline for scene-level perception in embodied intelligence. The source code and benchmark dataset will be made publicly available at https://github.com/GL-ZHU925/PanoAffordanceNet.

Problem

• Holistic Affordance Grounding
• 360-degree Indoor Environments
• Equirectangular Projection
• Semantic Dispersion
• Cross-scale Alignment

Innovation

• Holistic Affordance Grounding
• 360° Panoramic Perception
• Distortion-Aware Spectral Modulator
• Omni-Spherical Densification
• Embodied Intelligence