Scnet is a vehicle color classification network based on scloss, which can be easily combined with different backbone networks.

The Vehicle Color recognition Dataset contains 15601 vehicle images in eight colors, which are black, blue, cyan, gray, green, red, white and yellow. The images are taken in the frontal view captured by a high-definition camera with the resolution of 1920×1080 on the urban road. The collected data set is very challenging due to the noise caused by illumination variation, haze, and over exposure.Datasets in the OfRecord format are provided.

pip3 install -r requirements.txt --user

bash train.sh

The pretrained model on hustcolor.

bash infer.sh