By Lie Guo, Dongxing Wang, Linhui Li
With the development of deep learning, the performance of object detection has made great progress. But there are still some challenging problems, such as the detection accuracy of small objects and the efficiency of the detector. This paper proposes an accurate and fast single shot multibox detector (AFSSD) , which includes context comprehensive enhancement module (CCE) module and feature enhancement module (FEM).
- By default, we assume you have downloaded the file in the
AFSSD/weights
dir:
mkdir weights
cd weights
wget https://s3.amazonaws.com/amdegroot-models/vgg16_reducedfc.pth
- To train AFSSD using the train script simply specify the parameters listed in
train_AFSSD.py
as a flag or manually change them.
python train_AFSSD.py -d VOC -v AFSSD -s 320
To evaluate a trained network:
python test_AFSSD.py -d VOC -v AFSSD -s 320
Thanks for ruinmessi‘s project