Skip to main content
. 2016 Jun 21;6:28073. doi: 10.1038/srep28073

Figure 7. Snapshots of BSS results using movies.

Figure 7

Top: Four original images as hidden signal sources. Middle: Four superposed images provided as input to the model. Bottom: The final states of the outputs of the neural network reconstructed the original movies well (Supplementary Movie 2). We retrieved these movies from MotionElements (https://www.motionelements.com) and processed them accordingly.