Integrating contrastive cross-modal attention and stacked GRU for hand function rehabilitation robot control
Fig 6
The figure shows two confusion matrices, representing the model’s action classification results on the Ninapro DB5 dataset (left) and the MUSED-I dataset (right).