Frequency Guidance Matters: Skeletal Action Recognition by Frequency-Aware Mixed Transformer
This work addresses the problem of distinguishing similar actions in skeleton-based recognition for applications like human-computer interaction, though it is incremental in improving transformer-based approaches.
The paper tackles the challenge of recognizing similar skeletal actions with subtle motion differences by introducing a Frequency-aware Mixed Transformer (FreqMixFormer), which outperforms state-of-the-art methods on three popular datasets, including NTU RGB+D and NW-UCLA.
Recently, transformers have demonstrated great potential for modeling long-term dependencies from skeleton sequences and thereby gained ever-increasing attention in skeleton action recognition. However, the existing transformer-based approaches heavily rely on the naive attention mechanism for capturing the spatiotemporal features, which falls short in learning discriminative representations that exhibit similar motion patterns. To address this challenge, we introduce the Frequency-aware Mixed Transformer (FreqMixFormer), specifically designed for recognizing similar skeletal actions with subtle discriminative motions. First, we introduce a frequency-aware attention module to unweave skeleton frequency representations by embedding joint features into frequency attention maps, aiming to distinguish the discriminative movements based on their frequency coefficients. Subsequently, we develop a mixed transformer architecture to incorporate spatial features with frequency features to model the comprehensive frequency-spatial patterns. Additionally, a temporal transformer is proposed to extract the global correlations across frames. Extensive experiments show that FreqMiXFormer outperforms SOTA on 3 popular skeleton action recognition datasets, including NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets.