Auto Deep Compression by Reinforcement Learning Based Actor-Critic Structure
This addresses the problem of efficient model compression for resource-constrained applications, offering an automated approach that improves upon manual methods.
The paper tackles automated neural network compression by proposing Auto Deep Compression (ADC) using reinforcement learning, achieving a 4-fold reduction in FLOPs and 2.8% higher accuracy than manual compression for VGG-16 on ImageNet.
Model-based compression is an effective, facilitating, and expanded model of neural network models with limited computing and low power. However, conventional models of compression techniques utilize crafted features [2,3,12] and explore specialized areas for exploration and design of large spaces in terms of size, speed, and accuracy, which usually have returns Less and time is up. This paper will effectively analyze deep auto compression (ADC) and reinforcement learning strength in an effective sample and space design, and improve the compression quality of the model. The results of compression of the advanced model are obtained without any human effort and in a completely automated way. With a 4- fold reduction in FLOP, the accuracy of 2.8% is higher than the manual compression model for VGG-16 in ImageNet.