OrthoNets : Orthogonal Channel Attention Networks

PyTorch's implementation of the paper "OrthoNet : Orthogonal Channel Attention Networks".

Install

1. Install Linux
2. Install Anaconda
3. Install CUDA 11.X
4. Run the following comands in OrthoNet

conda env create --file orthonet.yaml
conda activate orthonet
pip install --extra-index-url https://developer.download.nvidia.com/compute/redist nvidia-dali-cuda110

5. Edit settings.sh if necessary.

Dataset acquisition and installation

If you are installing these datasets for the first time, place them in OrthoNet/datasets. Otherwise, please edit settings.sh to set your dataset location.

Dataset Locations

ImageNet-1000

Visit https://www.image-net.org/ and download ImageNet 2014 Dataset.

MS-COCO

Train     Images      : http://images.cocodataset.org/zips/train2017.zip
Val       Images      : http://images.cocodataset.org/zips/val2017.zip
Train/Val Annotations : http://images.cocodataset.org/annotations/annotations_trainval2017.zip

Birds

Train/Val Images : https://www.kaggle.com/datasets/gpiosenka/100-bird-species/download?datasetVersionNumber=59

Places365

Train/Val Images : https://www.kaggle.com/datasets/benjaminkz/places365/download?datasetVersionNumber=1

Setting up the models

The models can be found at

https://drive.google.com/drive/folders/1N_L7Cy5I2lwVTnjjYan2Z0rXNVw8rN-c?usp=share_link

Download the "pretrained_models" folder and place it in OrthoNet. Typically, the files are downloaded in multiple zips. Verify all 14 models are present.

Training and Testing

Training and testing the models is done by the run.sh script. The specific commands are found below.

Testing Pretrained Models

Method	Backbone	Dataset	Command
OrthoNet	ResNet-34	ImageNet	./run.sh orthonet_34_imagenet
OrthoNet	ResNet-50	ImageNet	./run.sh orthonet_50_imagenet
OrthoNet-MOD	ResNet-50	ImageNet	./run.sh orthonet_mod_50_imagenet
OrthoNet-MOD	ResNet-50	Birds	./run.sh orthonet_mod_50_birds
OrthoNet-MOD	ResNet-50	Places	./run.sh orthonet_mod_50_places
FcaNet	ResNet-50	Birds	./run.sh fcanet_50_birds
FcaNet	ResNet-50	Places	./run.sh fcanet_50_places
OrthoNet	ResNet-101	ImageNet	./run.sh orthonet_101_imagenet
OrthoNet-MOD	ResNet-101	ImageNet	./run.sh orthonet_mod_101_imagenet
OrthoNet-MOD	ResNet-50/FasterRCNN	COCO	./run.sh orthonet_mod_50_coco_faster_rcnn
OrthoNet-MOD	ResNet-101/FasterRCNN	COCO	./run.sh orthonet_mod_101_coco_faster_rcnn
FcaNet	ResNet-50/MaskRCNN	COCO	./run.sh fcanet_50_coco_mask_rcnn
OrthoNet-MOD	ResNet-50/MaskRCNN	COCO	./run.sh orthonet_mod_50_coco_mask_rcnn

Training Your Own Models

During training, FasterRCNN and MaskRCNN use the pretrained OrthoNet/FcaNet from the pretrained models folder.

Method	Backbone	Dataset	Command
OrthoNet	ResNet-34	ImageNet	./run.sh orthonet_34_imagenet train
OrthoNet	ResNet-50	ImageNet	./run.sh orthonet_50_imagenet train
OrthoNet-MOD	ResNet-50	ImageNet	./run.sh orthonet_mod_50_imagenet train
OrthoNet-MOD	ResNet-50	Birds	./run.sh orthonet_mod_50_birds train
OrthoNet-MOD	ResNet-50	Places	./run.sh orthonet_mod_50_places train
FcaNet	ResNet-50	Birds	./run.sh fcanet_50_birds train
FcaNet	ResNet-50	Places	./run.sh fcanet_50_places train
OrthoNet	ResNet-101	ImageNet	./run.sh orthonet_101_imagenet train
OrthoNet-MOD	ResNet-101	ImageNet	./run.sh orthonet_mod_101_imagenet train
OrthoNet-MOD	ResNet-50/FasterRCNN	COCO	./run.sh orthonet_mod_50_coco_faster_rcnn train
OrthoNet-MOD	ResNet-101/FasterRCNN	COCO	./run.sh orthonet_mod_101_coco_faster_rcnn train
FcaNet	ResNet-50/MaskRCNN	COCO	./run.sh fcanet_50_coco_mask_rcnn train
OrthoNet-MOD	ResNet-50/MaskRCNN	COCO	./run.sh orthonet_mod_50_coco_mask_rcnn train

Testing Your Trained Models

Method	Backbone	Dataset	Command
OrthoNet	ResNet-34	ImageNet	./run.sh orthonet_34_imagenet test
OrthoNet	ResNet-50	ImageNet	./run.sh orthonet_50_imagenet test
OrthoNet-MOD	ResNet-50	ImageNet	./run.sh orthonet_mod_50_imagenet test
OrthoNet-MOD	ResNet-50	Birds	./run.sh orthonet_mod_50_birds test
OrthoNet-MOD	ResNet-50	Places	./run.sh orthonet_mod_50_places test
FcaNet	ResNet-50	Birds	./run.sh fcanet_50_birds test
FcaNet	ResNet-50	Places	./run.sh fcanet_50_places test
OrthoNet	ResNet-101	ImageNet	./run.sh orthonet_101_imagenet test
OrthoNet-MOD	ResNet-101	ImageNet	./run.sh orthonet_mod_101_imagenet test
OrthoNet-MOD	ResNet-50/FasterRCNN	COCO	./run.sh orthonet_mod_50_coco_faster_rcnn test
OrthoNet-MOD	ResNet-101/FasterRCNN	COCO	./run.sh orthonet_mod_101_coco_faster_rcnn test
FcaNet	ResNet-50/MaskRCNN	COCO	./run.sh fcanet_50_coco_mask_rcnn test
OrthoNet-MOD	ResNet-50/MaskRCNN	COCO	./run.sh orthonet_mod_50_coco_mask_rcnn test