GraphiContact: Pose-aware Human-Scene Robust Contact Perception for Interactive Systems

2026-03-19

AI Summary

Existing monocular image-based methods for vertex-level human-scene contact prediction lack robustness to occlusion and perceptual noise, limiting their applicability in interactive systems. This work proposes GraphiContact, a framework that jointly performs 3D human mesh reconstruction and contact prediction by leveraging reconstructed geometry as a structural scaffold for reasoning and incorporating a pose-aware mechanism to enhance accuracy. The approach integrates human priors from two pretrained Transformer encoders, introduces a Single-Image Multi-uncertainty (SIMU) training strategy to simulate occlusion and noise during learning, and employs token-level adaptive routing to enable efficient single-branch inference at test time. Evaluated across five benchmark datasets, the method consistently achieves performance gains in both contact prediction and 3D human reconstruction tasks.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Computer Vision: Visual Reasoning & Symbolic Representations
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Monocular vertex-level human-scene contact prediction is a fundamental capability for interactive systems such as assistive monitoring, embodied AI, and rehabilitation analysis. In this work, we study this task jointly with single-image 3D human mesh reconstruction, using reconstructed body geometry as a scaffold for contact reasoning. Existing approaches either focus on contact prediction without sufficiently exploiting explicit 3D human priors, or emphasize pose/mesh reconstruction without directly optimizing robust vertex-level contact inference under occlusion and perceptual noise. To address this gap, we propose GraphiContact, a pose-aware framework that transfers complementary human priors from two pretrained Transformer encoders and predicts per-vertex human-scene contact on the reconstructed mesh. To improve robustness in real-world scenarios, we further introduce a Single-Image Multi-Infer Uncertainty (SIMU) training strategy with token-level adaptive routing, which simulates occlusion and noisy observations during training while preserving efficient single-branch inference at test time. Experiments on five benchmark datasets show that GraphiContact achieves consistent gains on both contact prediction and 3D human reconstruction. Our code, based on the GraphiContact method, provides comprehensive 3D human reconstruction and interaction analysis, and will be publicly available at https://github.com/Aveiro-Lin/GraphiContact.

Problem

• human-scene contact
• 3D human mesh reconstruction
• occlusion
• perceptual noise
• monocular vision

Innovation

• pose-aware contact perception
• 3D human mesh reconstruction
• Transformer priors
• vertex-level contact prediction
• uncertainty-aware training