SPECTRE: Spectral Pre-training Embeddings with Cylindrical Temporal Rotary Position Encoding for Fine-Grained sEMG-Based Movement Decoding

2025-12-27

AI Summary

To address the poor robustness of fine-grained motion decoding from surface electromyography (sEMG) signals—caused by their strong non-stationarity and low signal-to-noise ratio—this paper proposes SPECTRE, a domain-specific self-supervised framework. Methodologically, it integrates short-time Fourier transform (STFT), spectral clustering, masked modeling, and Transformer architecture within a self-supervised pretraining followed by supervised fine-tuning paradigm. Key innovations include: (1) a novel pseudo-labeling scheme via STFT spectrogram clustering and a spectral-domain masked prediction pretraining task; and (2) cylindrical rotational positional encoding (CyRoPE), explicitly modeling the circular topology of forearm electrode arrays and temporal dynamics. Evaluated on multi-source sEMG datasets—including real data from amputees—SPECTRE achieves new state-of-the-art performance in motion decoding, significantly outperforming both supervised baselines and generic self-supervised methods. Ablation studies confirm the critical contributions of spectral pretraining and CyRoPE.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Computer Vision: Motion & Tracking

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Decoding fine-grained movement from non-invasive surface Electromyography (sEMG) is a challenge for prosthetic control due to signal non-stationarity and low signal-to-noise ratios. Generic self-supervised learning (SSL) frameworks often yield suboptimal results on sEMG as they attempt to reconstruct noisy raw signals and lack the inductive bias to model the cylindrical topology of electrode arrays. To overcome these limitations, we introduce SPECTRE, a domain-specific SSL framework. SPECTRE features two primary contributions: a physiologically-grounded pre-training task and a novel positional encoding. The pre-training involves masked prediction of discrete pseudo-labels from clustered Short-Time Fourier Transform (STFT) representations, compelling the model to learn robust, physiologically relevant frequency patterns. Additionally, our Cylindrical Rotary Position Embedding (CyRoPE) factorizes embeddings along linear temporal and annular spatial dimensions, explicitly modeling the forearm sensor topology to capture muscle synergies. Evaluations on multiple datasets, including challenging data from individuals with amputation, demonstrate that SPECTRE establishes a new state-of-the-art for movement decoding, significantly outperforming both supervised baselines and generic SSL approaches. Ablation studies validate the critical roles of both spectral pre-training and CyRoPE. SPECTRE provides a robust foundation for practical myoelectric interfaces capable of handling real-world sEMG complexities.

Problem

• Decoding fine-grained movement from noisy sEMG signals
• Addressing suboptimal SSL due to lack of physiological bias
• Modeling cylindrical topology of forearm electrode arrays

Innovation

• Spectral pre-training via masked STFT pseudo-label prediction
• Cylindrical rotary position encoding for sensor topology
• Domain-specific SSL framework for robust sEMG decoding