fig3

Figure 3. Gesture capture, mapping and recognition system. [A(i)] Schematic of the gesture capture and mapping system; [A(ii)] Demonstration of ten common sign-language gestures; (B) Real-time signals generated by the smart glove for the ten representative gestures; [C(i)] Architecture of the CNN model; [C(ii)] Confusion matrix of the test set. Some image elements are photographs taken by the authors or obtained from open-source websites (Pixabay and FreeImages). CNN: Convolutional neural network; MCU: microcontroller unit.