A Novel Deep Learning Approach for Recognizing Stereotypical Motor Movements within and across Subjects on the Autism Spectrum Disorder

Figure 4

Plots ((a.1)-(a.2)) and ((b.1)-(b.2)) correspond to time and frequency-domain acceleration signal samples, respectively, which are extracted from the “SMM dataset”. The red, green, and blue curves represent x, y, and z signals, respectively. Samples ((a.1)-(a.2)) and ((b.1)-(b.2)) have lengths of 90 (~ 1 second) and 50 (~50 points interval representing frequency in the range of 0-3Hz), respectively.