Object Detection using Single Shot MultiBox Detector with Caffe MobileNet on OpenCV in Python.
Single Shot MultiBox Detectors can be divided into two parts:
- Extracting Features using a base network
- Using Convolution Filters to make predictions
This implementation makes use of the MobileNet deep learning CNN architecture as the base network.
Caffe is a deep learning framework developed by the Berkely AI Research and Community Contributors. Caffe repo. It is a much faster way of training images with over 6 million images per day using an Nvidia K-40 GPU
python detectDNN.py -p Caffe/SSD_MobileNet_prototxt -m Caffe/SSD_MobileNet.caffemodel