入门DeepLearning-小乐陪你读paper

2016-11-09 安兴乐-siler 更多博文 » 博客 » GitHub »

神经网络

原文链接 http://www.lelovepan.cn/2016/11/09/DeepLearning-State-of-Art%E8%AE%BA%E6%96%87%E9%9B%86.html
注：以下为加速网络访问所做的原文缓存，经过重新格式化，可能存在格式方面的问题，或偶有遗漏信息，请以原文为准。

Text Recognition

  • Fully Convolutional Recurrent Network for Handwritten Chinese Text Recognition paper
  • Deep LSTM Networks for Online Chinese Handwriting Recognition 2016 ICFHR
  • Convolution Multi-directional Recurrent Network for Offline Handwritten Text Recognition 2016 ICFHR
  • Sequence to Sequence Learning with Neural Networks paper
  • Deep Neural Network based Hidden Markov Model for Offline Handwritten Chinese Text Recognition paper
  • Scene Text Script Identification with Convolutional Recurrent Neural Networks paper
  • A Simple Way to Initialize Recurrent Networks of Rectified Linear Units paper Google paper，强烈推荐
  • Generating Sequences With Recurrent Neural Networks paper Alex论文，RNN入门首选
  • Recurrent Neural Network Regularization paper 就是说一下怎么在RNN中应用Dropout
  • Recurrent Batch Normalization paper RNN中应用normalization效果好

Style Transfer

  • A learned representation for artistic style, okt 2016, Google, arxiv, demo
  • Fast Style Transfer in TensorFlow, github 使用Tensorflow来做，最好提前下好VGG模型
  • https://arxiv.org/abs/1508.06576
  • https://arxiv.org/abs/1607.08022
  • https://cs.stanford.edu/people/jcjohns/eccv16/

Object Recognition

  • Learning to Make Better Mistakes: Semantics-aware Visual Food Recognition, okt 2016, IBM, paper
  • T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos, aug 2016, github, arxiv
  • Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning, aug 2016, Google, arxiv
  • Residual Networks of Residual Networks: Multilevel Residual Networks, aug 2016, arxiv
  • Training Region-based Object Detectors with Online Hard Example Mining, apr 2016, Facebook, arxiv
  • Deep Residual Learning for Image Recognition, dec 2015, arxiv
  • SSD: Single Shot MultiBox Detector, dec 2015, Google, github, arxiv state-of-the-art
  • ParseNet: Looking Wider to See Better, jun 2015, arxiv
  • You Only Look Once: Unified, Real-Time Object Detection, jun 2015, Facebook, arxiv
  • Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks, jun 2015, Microsoft/Facebook arxiv state-of-the-art
  • Selective Search for Object Recognition, 2012, paper
  • Rich feature hierarchies for accurate object detection and semantic segmentation, 2014, paper