1. It is better to use https://github.com/lgvaz/mantisshrimp
2. All the features in this (plus extra features) have been added to mantisshrimp.
3. It is a complete framework for object detection and more scalable.

Faster RCNN Fine-Tune Implementation in Pytorch.

1. git clone the repo

git clone https://github.com/oke-aditya/pytorch_fasterrcnn.git

2. install the requirements (will add)

pip install -r requirements.txt

Simply edit the config file to set your hyper parameters.

3. Keep the training and validation csv file as follows

NOTE

Do not use target as 0 class. It is reserved as background.

image_id xtl ytl xbr ybr      target
1        xmin ymin xmax ymax   1
1        xmin ymin xmax ymax   2

2		 xmin ymin xmax ymax   3

4. Simply edit the config file to set your hyper parameters

5. Run the train.py file

• It works for multiple class object detection.

• Note that backbones are pretrained on imagenet.

• Following backbones are supported

1. vgg11, vgg13, vgg16, vgg19
2. resnet18, resnet34, resnet50, resnet101, resnet152
3. renext101
4. mobilenet_v2

Sample Outputs

If you like the implemenation or have taken an inspiration do give a star :-)