Impact of Disentanglement on Pruning Neural Networks

Efficient model compression techniques are required to deploy deep neural networks (DNNs) on edge devices for task specific objectives. A variational autoencoder (VAE) framework is combined with a pruning criterion to investigate the impact of having the network learn disentangled representations on the pruning process for the classification task.

Poster from the Computer Vision, Imaging, and Machine Intelligence Research Group (CVI²) at SnT, University of Luxembourg. Selected for poster presentation during the first edition of the International Symposium on Computational Sensing ISCS23 in Luxembourg.