Towards a Segmentation Solution: Hidden Markov Models for Sleep-Wake Segmentation
What can we learn about transitioning between sleep and wake states for longer full-video segmentation on the edge?
Extracting Pose and Optical Flow features
NN-HMM architecture
