Multi-Modal Graph Convolutional Network with Sinusoidal Encoding for Robust Human Action Segmentation

TL;DR

This paper introduces a multi-modal graph convolutional network with sinusoidal encoding and a novel data augmentation technique to improve the accuracy and temporal coherence of human action segmentation, especially under noisy conditions.

Contribution

The paper presents a novel multi-modal GCN framework with sinusoidal encoding, hierarchical feature fusion, and SmoothLabelMix augmentation for robust human action segmentation.

Findings

Achieves state-of-the-art segmentation accuracy on Bimanual Actions Dataset.

Effectively reduces over-segmentation errors in noisy data.

Demonstrates robustness to low-frame-rate visual data.

Abstract

Accurate temporal segmentation of human actions is critical for intelligent robots in collaborative settings, where a precise understanding of sub-activity labels and their temporal structure is essential. However, the inherent noise in both human pose estimation and object detection often leads to over-segmentation errors, disrupting the coherence of action sequences. To address this, we propose a Multi-Modal Graph Convolutional Network (MMGCN) that integrates low-frame-rate (e.g., 1 fps) visual data with high-frame-rate (e.g., 30 fps) motion data (skeleton and object detections) to mitigate fragmentation. Our framework introduces three key contributions. First, a sinusoidal encoding strategy that maps 3D skeleton coordinates into a continuous sin-cos space to enhance spatial representation robustness. Second, a temporal graph fusion module that aligns multi-modal inputs with differing resolutions via hierarchical feature aggregation, Third, inspired by the smooth transitions inherent to human actions, we design SmoothLabelMix, a data augmentation technique that mixes input sequences and labels to generate synthetic training examples with gradual action transitions, enhancing temporal consistency in predictions and reducing over-segmentation artifacts. Extensive experiments on the Bimanual Actions Dataset, a public benchmark for human-object interaction understanding, demonstrate that our approach outperforms state-of-the-art methods, especially in action segmentation accuracy, achieving F1@10: 94.5% and F1@25: 92.8%.