论文笔记——N2N Learning: Network to Network Compression via Policy Gradient Reinforcement Learning
2017-11-19 20:28
701 查看
论文地址：https://arxiv.org/abs/1709.06030
1. 论文思想
利用强化学习，对网络进行裁剪，从Layer Removal和Layer Shrinkage两个维度进行裁剪。 一个是对层判断是否进行裁剪，一个是判断一层中的参数的裁剪。2. 原理图
3. 实现细节
将层信息进行编码表示，然后送入双端的LSTM中，最后通过Softmax学出多个行为的概率。然后来决定层的裁剪信息。4. 结果
ResNet-34上实现了10倍的压缩。
内容来自用户分享和网络整理，不保证内容的准确性，如有侵权内容，可联系管理员处理 
相关文章推荐
- 论文笔记之：Dueling Network Architectures for Deep Reinforcement Learning
- 论文解读之 One-to-Many Network for Visually Pleasing Compression Artifacts Reduction
- 论文笔记：Learning to Detect Violent Videos using Convolutional Long Short-Term Memory
- 《reinforcement learning：an introduction》第十三章《Policy Gradient Methods》总结
- Learning to learn by gradient descent by gradient descent 笔记
- 论文笔记之：Heterogeneous Image Features Integration via Multi-Modal Semi-Supervised Learning Model
- OpenCV编程实现LeCun论文（Gradient-Based Learning Applied to Document Recognition）中的CNN
- 【论文笔记】Sequence to sequence Learning with Neural Networks
- 基于2-channel network的图片相似度判别_2015Learning to Compare Image Patches via Convolutional Neural Networks
- 深度学习论文笔记（七）---Deconvolution network-2016年（Learning Deconvolution Network for Semantic Segmentation ）
- 【学习笔记】WEEK 1_Introduction to Deep Learning_What is a neural network?
- 论文笔记：Knowing When to Look: Adaptive Attention via A Visual Sentinel for Image Captioning
- 论文笔记之：Playing Atari with Deep Reinforcement Learning
- 论文笔记：Label Refinement Network for Coarse-to-Fine Semantic Segmentation
- Reinforce learning: Lecture 1 Introduction to Reinforcement Learning 学习笔记
- 论文笔记：An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application
- 【论文笔记】From Facial Parts Responses to Face Detection: A Deep Learning Approach
- 论文笔记之：DualGAN: Unsupervised Dual Learning for Image-to-Image Translation
- silver课程笔记----Lecture 1: Introduction to Reinforcement Learning