An off the shelf implementation of a single shot detector which uses MobileNet-v2 instead of VGG-16 for feature extraction.
The model is expected to be lightweight and deployable in embedded systems mainly due to a MobileNet-v2 feature extractor.
Applications and the working of this model will be added in the future.
To inference in real time using this model:
python3 run_ssd_live_demo.py