Pedestrian-RCNN

This code extends py-faster-rcnn by adding:

• ResNet implementation.

• Online Hard Example Mining.

• Caltech Dataset train and test interface.

• Add RPN tools.

Base

• The faster rcnn code is based on py-faster-rcnn.

• The ohem code is based on ohem.

• To reduce the memory usage, we use batchnorm layer in Microsoft's caffe

• The caltech dataset interface based on GBJim-py-faster-rcnn

Installation

The installation and useage are same as Faster R-CNN

1. clone the repository and caffe submodule

2. build lib and caffe

3. train your model command like: ./experiments/scripts/faster_rcnn_end2end.sh 0 ResNet50 caltech_reasonable

Notices

For different datasets, the interval frame rate is different, so it need to be changed.

In ./lib/datasets/caltech.py, change the start_frame and frame_rate of insert_frame function.