 This study investigated the use of a wearable, flexible multi-channel capacitive wristband to improve the accuracy of hand gesture recognition over several days of usage. The results showed that the triplet network outperformed other methods such as a convolutional neural network with softmax cross entropy loss. Additionally, the size of the capacitive array had an effect on the overall accuracy of the system. This article was authored by Tianan Wang, Yubiao Zhao, and Chenying Wang.