Prune Your Neurons Blindly: Neural Network Compression through Structured Class-blind Pruning
| Autor: | Abdullah Salama, Tassilo Klein, Oleksiy Ostapenko, Moin Nabi |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Artificial neural network
Computer science business.industry Deep learning 02 engineering and technology 010501 environmental sciences 01 natural sciences Redundancy (information theory) Computer engineering Compression (functional analysis) 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Limit (mathematics) Artificial intelligence Pruning (decision trees) business 0105 earth and related environmental sciences |
| Zdroj: | ICASSP |
| Popis: | High performance of deep learning models typically comes at cost of considerable model size and computation time. These factors limit applicability for deployment on memory and battery constrained devices such as mobile phones or embedded systems. In this work, we propose a novel pruning technique that eliminates entire filters and neurons according to their relative L1-norm as compared to the rest of the network, yielding more compression and decreased parameters’ redundancy. The resulting network is non-sparse, however, much more compact and requires no special infrastructure for deployment. We prove the viability of our method by achieving 97.4%, 86.1%, 47.8% and 53% compression of LeNet-5, VGG-16, ResNet-56 and ResNet-110 respectively, exceeding state-of-the-art compression results reported on VGG-16 and ResNet without losing any performance compared to the baseline. Our approach does not only exhibit good performance but is also easy to implement. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|