Abstract: Wearable human activity recognition (HAR) has greatly benefited from deep convolutional neural networks. However, while increasing network width or depth improves accuracy, it also leads to ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results