YOLOv3, RetinaNet, Faster RCNNs et al.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
images
LICENSE
README.md
dataset.sh
obj_detection_single.py
object_detection_multi.py

README.md

Object-detection

Requirements

  • Pytorch >= 0.3

  • opencv-python (for plotting purpose)

  • fastai library

     To install it
     $ pip install https://github.com/fastai/fastai/archive/master.zip')
    
  • torchvision

  • Some extra libs : $apt update && apt install -y libsm6 libxext6 (May vary for other Linux distro)

Dataset

  • Pacal VOC 2007 & 2012

There are two variants available. I made of 2007 version but you can expect better results with 2012 (more computation) To download and manage directory run sh dataset.sh

Training

$ python obj_detection_single.py

By deafult your model will be saved in models directory. Create one if it does not exist and set the path as per your directory structure.

Results

Labelling

p1

Bounding boxes

p2