Goal driven network pruning for object recognition
| Autor: | Abdullah Bulbul, Cagri Kaplan |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
business.industry
Computer science Deep learning Cognitive neuroscience of visual object recognition 02 engineering and technology Machine learning computer.software_genre 01 natural sciences Task (project management) Artificial Intelligence 0103 physical sciences Signal Processing Human visual system model 0202 electrical engineering electronic engineering information engineering Benchmark (computing) 020201 artificial intelligence & image processing Computer Vision and Pattern Recognition Artificial intelligence Pruning (decision trees) Layer (object-oriented design) 010306 general physics business computer Software MNIST database |
| Zdroj: | Pattern Recognition. 110:107468 |
| ISSN: | 0031-3203 |
| DOI: | 10.1016/j.patcog.2020.107468 |
| Popis: | Pruning studies up to date focused on uncovering a smaller network by removing redundant units, and fine-tuning to compensate accuracy drop as a result. In this study, unlike the others, we propose an approach to uncover a smaller network that is competent only in a specific task, similar to top-down attention mechanism in human visual system. This approach doesn't require fine-tuning and is proposed as a fast and effective alternative of training from scratch when the network focuses on a specific task in the dataset. Pruning starts from the output and proceeds towards the input by computing neuron importance scores in each layer and propagating them to the preceding layer. In the meantime, neurons determined as worthless are pruned. We applied our approach on three benchmark datasets: MNIST, CIFAR-10 and ImageNet. The results demonstrate that the proposed pruning method typically reduces computational units and storage without harming accuracy significantly. (c) 2020 Elsevier Ltd. All rights reserved. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect