A background subtraction library
Boosting Convolutional Neural Networks
Caffe with heatmap regression & spatial fusion layers. Useful for any CNN image position regression task.
Caffe models (imagenet pretrain) and prototxt generator scripts for inception_v3 \ inception_v4 \ inception_resnet \ fractalnet \ resnext
MatConvNet and Caffe repo with compact bilinear and bilinear pooling functionality added
An independent implementation of "CrowdNet: A Deep Convolutional Network for Dense Crowd Counting"
Caffe code for Densely Connected Convolutional Networks (DenseNets)
Detection Evaluation with AP criterion
Code that accompanies my blog post outlining five video classification methods in Keras and TensorFlow
Repo for the course IFT6266
collections of papers and projects
一个拍照做题程序。输入一张包含数学计算题的图片，输出识别出的数学计算式以及计算结果。This is a mathematic expression recognition project.
Compare the pedestrain detection using Faster RCNN and RPN+BF
Faster R-CNN (Python implementation) -- see https://github.com/ShaoqingRen/faster_rcnn for the official MATLAB version
A text replacement based faster rcnn data interface generator
faster rcnn Online hard example mining
pyNetBuilder is a modular pytonic interface with builtin modules for generating popular caffe prototxt network file definitions.
image captions generation with spatial and channel-wise attention
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters
Video Classification using 2 stream CNN
Weighted Softmax Loss Layer for Caffe