Abstract: Convolutional Neural Networks (CNNs) require costly hardware for real-time processing due to massive multiplications. In video-based inference, many pixels remain unchanged between frames.
Some results have been hidden because they may be inaccessible to you
Show inaccessible results