Pruning and Quantization for Deep Neural Network Acceleration: A Survey

no code implementations • 24 Jan 2021 • Tailin Liang, John Glossner, Lei Wang, Shaobo Shi, Xiaotong Zhang

We discuss trade-offs in element-wise, channel-wise, shape-wise, filter-wise, layer-wise and even network-wise pruning.

Quantization

Dynamic Runtime Feature Map Pruning

1 code implementation • 24 Dec 2018 • Tailin Liang, Lei Wang, Shaobo Shi, John Glossner

Of the networks considered, those using ReLU (AlexNet, SqueezeNet, VGG16) contain a high percentage of 0-valued parameters and can be statically pruned.