目標檢測
OverFeat
文章：《OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks》
rcnn
代碼：https://github.com/rbgirshick/rcnn，prototxt文件在\rcnn-master\finetuning\voc_2012_prototxt
fast-rcnn
文章：《Fast R-CNN》
代碼：https://github.com/rbgirshick/fast-rcnn，prototxt文件在.\fast-rcnn-master\models\CaffeNet
faster-rcnn
文章：《Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks》
代碼：https://github.com/rbgirshick/py-faster-rcnn，prototxt文件在.\py-faster-rcnn-master\models\pascal_voc\VGG16\faster_rcnn_end2end
tensorflow的代碼：tf-faster-rcnn代碼理解
YOLO
文章：《You Only Look Once: Unified, Real-Time Object Detection》
代碼：https://github.com/xingwangsfu/caffe-yolo
文章：《YOLO9000: Better, Faster, Stronger》
SSD
文章：《SSD: Single Shot MultiBox Detector》
代碼：https://github.com/weiliu89/caffe/tree/ssd
由於SSD中需要運行python examples/ssd/ssd_pascal.py來生成solver.prototxt, train.prototxt, test.prototxt, deploy.prototxt。如果你沒有硬件運行代碼或嫌麻煩，可以在這裏直接看https://gist.github.com/JeffOwOSun/1c284e6177c38b17d83da1db7b8c8ce7
FPN
文章：《Feature Pyramid Networks for Object Detection》
mask rcnn
文章：《Mask R-CNN》
語義分割
FCN
文章：《Fully Convolutional Networks for Semantic Segmentation》
代碼：https://github.com/shekkizh/FCN.tensorflow
網絡結構
VGG
文章：《Very Deep Convolutional Networks for Large-scale image recognition》
googleNet
文章：
Inception v1:《Going Deeper with Convolutions》
Inception v2:《Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate 》
Inception v3:《Rethinking the Inception Architecture for Computer Vision》
resnet
文章：《Deep Residual Learning for Image Recognition》
代碼：https://github.com/KaimingHe/deep-residual-networks
補充文章：《Residual Networks Behave Like Ensembles of Relatively Shallow Networks》
技巧
梯度遞減優化算法綜述
《An overview of gradient descent optimization algorithms*》
dropout防止過擬合
《Dropout: A Simple Way to Prevent Neural Networks from
Overfitting》