This thesis proposes a novel neural network architecture, the Convolutional Spider Neural Network (CS-Net), complemented by a transfer learning strategy, to classify hybrid gestures combining wrist postures and hand movements. The CS-Net framework integrates diverse surface electromyography (sEMG) features, including raw signals and fast Fourier transform (FFT) representations, through a multi-stream information fusion mechanism to enhance classification performance. The proposed transfer learning strategy involves pre-training the model on specific wrist postures and fine-tuning it on the full set of hybrid gestures, leveraging the intrinsic relationships between composite gestures and their constituent movements to improve accuracy.

This thesis evaluated the framework through extensi...[ Read more ]

This thesis proposes a novel neural network architecture, the Convolutional Spider Neural Network (CS-Net), complemented by a transfer learning strategy, to classify hybrid gestures combining wrist postures and hand movements. The CS-Net framework integrates diverse surface electromyography (sEMG) features, including raw signals and fast Fourier transform (FFT) representations, through a multi-stream information fusion mechanism to enhance classification performance. The proposed transfer learning strategy involves pre-training the model on specific wrist postures and fine-tuning it on the full set of hybrid gestures, leveraging the intrinsic relationships between composite gestures and their constituent movements to improve accuracy.

This thesis evaluated the framework through extensive offline experiments using a dataset of 12 hybrid gestures (combining three wrist postures and four hand movements) collected from six subjects, comparing its performance against three state-of-the-art deep learning algorithms. Additionally, this thesis validated its generalizability using 30 actions from the publicly available Ninapro dataset. To demonstrate practical applicability, this thesis conducted real-time online experiments involving object grasping tasks. The results show that CS-Net significantly improves sEMG classification accuracy, while the transfer learning strategy further enhances performance. Moreover, the algorithm achieved a high success rate in online experiments, confirming its robustness and practical utility for real-world applications.