ADDA.PyTorch-resnet

Implementation of "Adversarial Discriminative Domain Adaptation" in PyTorch.

This repo is mostly based on https://github.com/Fujiki-Nakamura/ADDA.PyTorch

Note

Before running the training code, make sure that DATASETDIR environment variable is set to dataset directory.

Environment

• Python 3.8.5
• PyTorch 1.6.0

Example

For training on SVHN-MNIST

$ python train_source.py --logdir outputs
$ python main.py --logdir outputs --trained outputs/best_model.pt --slope 0.2

For training on Office dataset using ResNet-50

$ python core/train_source_rn50.py --n_classes 31 --lr 1e-5 --src_cat amazon --tgt_cat webcam
$ python main.py --n_classes 31 --trained outputs/garbage/best_model.pt --lr 1e-5 --d_lr 1e-4 --logdir outputs --model resnet50 --src-cat amazon --tgt-cat webcam

Result

SVHN -> MNIST

	Paper	This Repo
Source only	0.601	0.646
ADDA	0.760	0.805

Office-31 Amazon -> Office-31 Webcam

	Paper	This Repo
Source only	0.684	0.714
ADDA	0.862	0.831

Resources

• https://arxiv.org/pdf/1702.05464.pdf
• https://github.com/Fujiki-Nakamura/ADDA.PyTorch
• https://github.com/erictzeng/adda/issues/11
• https://github.com/corenel/pytorch-adda/issues/15