- 01-lighting-hydra-mnist cnn
- 02-wheat (lighting, hydra) fasterrcnn_resnet50_fpn albumentations cv2
- 03-torchvision-finetuning maskrcnn_resnet50_fpn torchvision.transforms PIL
- 04-face-detection (lighting) FasterRCNN albumentations + custom skimage
- 05-effdet (lighting) effdet albumentations PIL
- lighting
- hydra
- jpg를 read할때 뭘 사용하는가?(cv2, PIL, sk-image, torchvision)
- transform을 하는 방법(직접, albumentation, torchvision)
- model의 input, output type