pytorch-ssd for TME Dataset

A Pytorch implementation of Single Shot MultoBox Detector. The original code can be found here.

   

Installation

For installation, you can follow this repository (https://github.com/amdegroot/ssd.pytorch) The only difference is that this code is perfectly operated with pytorch 0.4.0.

Modification

I added some Modules such as Extension Module and Spatial Weight Module for accurate vehicle detection.

\

TME Dataset

You can get the TME Dataset from here (http://cmp.felk.cvut.cz/data/motorway/).

Training

• First download the fc-reduced VGG-16 PyTorch base network weights at: https://s3.amazonaws.com/amdegroot-models/vgg16_reducedfc.pth
• By default, we assume you have downloaded the file in the ssd.pytorch/weights dir:

mkdir weights
cd weights
wget https://s3.amazonaws.com/amdegroot-models/vgg16_reducedfc.pth

• To train SSD_TME using the train script simply specify the parameters listed in train_tme_sw_deconv.py as a flag or manually change them.

python train_tme_sw_deconv.py