研究方法 CNN
准备资料
1 数据
1.1 小型分类数据准备(代码)
学习将一个数据集合转为LMDB
http://www.cnblogs.com/alexcai/p/5469436.html
1.2 密度数据准备(数据)
mall-dataset数据库下载链接
http://personal.ie.cuhk.edu.hk/~ccloy/downloads_mall_dataset.htmlz
http://blog.csdn.net/garfielder007/article/details/51519575
1.3 密度图生成(代码)
公开数据库给点的常常是点集,密度的使用是通过一个脉冲函数卷积高斯核模拟真实场景
密度图生成原理http://blog.csdn.net/snwang_miss/article/details/77851203
https://github.com/davideverona/deep-crowd-counting_crowdnet
1.4 训练数据准备(代码数据转换)
1.4.1 输入数据使用.h5格式存储
http://blog.csdn.net/shuzfan/article/details/52503683
caffe HDF5Data 层使用及数据生成
http://blog.csdn.net/u011762313/article/details/48851015
Caffe中HDF5Data例子
1.4.2 输入数据使用密度图
2 网络使用
网络
人群分析 Slicing Convolutional Neural Network for Crowd Video Understanding CVPR2016
http://www.ee.cuhk.edu.hk/~jshao/SCNN.html
Caffe code: https://github.com/amandajshao/Slicing-CNN
人体检测 End-to-end people detection in crowded scenes CVPR2016
https://github.com/Russell91/ReInspect
人群计数 Single-Image Crowd Counting via Multi-Column Convolutional Neural Network CVPR2016
https://github.com/svishwa/crowdcount-mcnn
https://github.com/leeyeehoo/Reduplication-of-Single-Image-Crowd-Counting-via-MCNN-on-UCF-Dataset
人群计数 Switching Convolutional Neural Network for Crowd Counting CVPR2017
https://github.com/val-iisc/crowd-counting-scnn
人群分析 Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction
https://www.microsoft.com/en-us/research/publication/deep-spatio-temporal-residual-networks-for-citywide-crowd-flows-prediction/
https://github.com/lucktroy/DeepST/tree/master/scripts/papers/AAAI17
人群分析 物体计数 Towards perspective-free object counting with deep learning ECCV2016
https://github.com/gramuah/ccnn
人群计数 CNN-based Cascaded Multi-task Learning of High-level Prior and Density Estimation for Crowd Counting AVSS 2017
https://github.com/svishwa/crowdcount-cascaded-mtl
人群计数 CrowdNet: A Deep Convolutional Network for Dense Crowd Counting Caffe:
https://github.com/davideverona/deep-crowd-counting_crowdnet
ubuntu+caffe
安装方法
http://blog.csdn.net/autocyz/article/details/52299889
http://blog.csdn.net/xuzhongxiong/article/details/52717285
3 运行
http://blog.csdn.net/qq_14845119/article/details/68946727#comments
http://blog.csdn.net/qq_14845119/article/details/68946693