no code implementations • 29 Sep 2021 • Łukasz Treszczotko, Pawel Kubik
The framework contains 1) an automatic channel detection algorithm that groups the interdependent blocks of channels; 2) a non-iterative pruning algorithm that learns channel importance directly from feature maps while masking the coupled computational blocks using Gumbel-Softmax sampling and 3) a hierarchical knowledge distillation approach to fine-tune the pruned neural networks.
no code implementations • 19 Apr 2020 • Marcin Możejko, Tomasz Latkowski, Łukasz Treszczotko, Michał Szafraniuk, Krzysztof Trojanowski
Recent advancements in Neural Architecture Search(NAS) resulted in finding new state-of-the-art Artificial Neural Network (ANN) solutions for tasks like image classification, object detection, or semantic segmentation without substantial human supervision.