Demo implementation of OSHOT: One SHOT unsupervised cross domain detection.

This code is based on Maskrcnn-benchmark and uses Pytorch and CUDA.

This readme will guide you through a full run of our method for the Pascal VOC -> AMD benchmarks. Configuration files are provided also to perform other experiments.

Installation

Check INSTALL.md for installation instructions.

Datasets

Create a folder named datasets and include VOC2007 and VOC2012 source datasets (download from Pascal VOC's website).

Download and extract clipart1k, comic2k and watercolor2k from authors' website.

Performing OSHOT pretraining

To perform the pretraing using Pascal VOC as source dataset:

python tools/train_net.py --config-file configs/amd/voc_pretrain.yaml

By default training and inference are performed on a single GPU.

The final model will be saved in VOC_RS_baseline/model_final.pth.

Testing pretrained model

You can test a pretrained model on one of the AMD referring to the correct config-file. For example for clipart:

python tools/test_net.py --config-file configs/amd/oshot_clipart_target.yaml --ckpt VOC_RS_baseline/model_final.pth

Performing OSHOT adaptation

To use OSHOT adaptation rocedure and obtain results on one of the AMD please refer to one of the config files. For example for clipart:

python tools/oshot_net.py --config-file configs/amd/oshot_clipart_target.yaml --ckpt VOC_RS_baseline/model_final.pth