This is the current structure of the project. Note that the project structure is likely to change in the future.
├── cfsa # package containing the proposed few shot algorithm
├── comparisons # other few-shot algorithms used for comparisons (when possible directly cloned from GitHub)
├── data # training and evaluation datasets
│ ├── coco
│ └── lvis
├── scripts # the only files that contains a main (training, evaluation, etc.)
└── utils
├── dataset # dataset utilities (dataset classes, data pre/post processing)
└── model # torch.nn.Modules that are not directly part of the algorithm (e.g. feature extractors)