ADC: Automated Deep Compression and Acceleration with Reinforcement Learning

By Yihui He, Song Han. 2018

77 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Liang-Chieh Chen
2. Quoc V. Le
3. Yi Yang
4. Chunhua Shen
5. Mingxing Tan
6. Dacheng Tao
7. Luc Van Gool
8. Jiahui Yu

Citations in the corpus

• Searching for MobileNetV3 (2019, methodology): “Focusing on adapting existing networks to constrained mobile platforms, [48, 15, 12] proposed more efficient automated network simplification algorithms.”
• NetAdapt: Platform-Aware Neural Network Adaptation for Mobile Applications (2018, methodology): “The results are shown and compared with the multipliers [9] and ADC [8] in Fig.”
• Importance Estimation for Neural Network Pruning (2019, methodology): “A variety of pruning methods have been proposed, based on greedy algorithms [26, 33], sparse regularization [21, 23, 32], and reinforcement learning [13].”
• Quantizing deep convolutional networks for efficient inference: A whitepaper (2018, methodology): “Reinforcement learning has been applied successfully towards this problem in [33].”
• Training Compact Neural Networks with Binary Weights and Low Precision Activations (2018, background): “However, accuracy is roughly directly proportional to log(FLOPs) [3, 10] and training accurate CNNs always requires billions of FLOPs.”
• LadaBERT: Lightweight Adaptation of BERT through Hybrid Model Compression (2020, methodology): “In (He & Han, 2018), an AutoML pipeline was adopted for model compression.”
• Evolutionary Compression of Deep Neural Networks for Biomedical Image Segmentation (2020, methodology): “Also, there are some automated pruning approaches [49], [50], which are capable of discovering the”
• Structured Probabilistic Pruning for Convolutional Neural Network Acceleration (2017, methodology): “Further, [12] employed Re-inforcement Learning to automatically learn the pruning ratio for different layers.”
• Meta Filter Pruning to Accelerate Deep Convolutional Neural Networks (2019, methodology): “[19, 20, 15, 21, 6, 29, 33, 31, 37, 16, 12] utilize the training data to determine the pruned filters.”
• Filter Pruning via Geometric Median for Deep Convolutional Neural Networks Acceleration (2018, methodology): “Some filter pruning approaches [23, 25, 16, 27, 7, 33, 39, 37, 46, 14, 18, 22] need to utilize training data to determine the pruned filters.”