Table 6.

Comparison with the state of the art approaches using UCI HAR dataset. The number in bold indicates the highest accuracy.

Approaches	Accuracy (%)
SVM [10]	96.37
DCNN+ [12]	97.59
FFT+Convnet [14]	95.75
TSCHMM [13]	96.74
CNN+Sharpen (Ours)	97.62