A PyTorch implementation of a DroNet Object Detector

This repository is based on pytorch-yolo-v3

Requirements

sudo -H pip3 install torch==0.4.0

Using PyTorch 0.3 will break the detector.

Clone, and cd into the repo directory.

python3 video_demo.py --cfg cfg/DroNet_car.cfg --weights cfg/DroNet_car.weights --video Car_Parking.mov --classes 1

python3 video_demo.py --cfg cfg/DroNetV3_car.cfg --weights cfg/DroNetV3_car.weights --video Car_Parking.mov --classes 2

python3 video_demo.py --cfg cfg/DroNetV3_car.cfg --weights cfg/DroNetV3_car.weights --video Car_Crossroad.mp4 --classes 2 --reso 1024

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
cfg		cfg
data		data
det		det
imgs		imgs
.gitignore		.gitignore
Car_Crossroad.mp4		Car_Crossroad.mp4
Car_Parking.mov		Car_Parking.mov
Car_Road.MOV		Car_Road.MOV
README.md		README.md
__init__.py		__init__.py
bbox.py		bbox.py
cam_demo.py		cam_demo.py
darknet.py		darknet.py
detect.py		detect.py
dog-cycle-car.png		dog-cycle-car.png
pallete		pallete
preprocess.py		preprocess.py
util.py		util.py
video_demo.py		video_demo.py
video_demo_half.py		video_demo_half.py