Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time read more
Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution (CVPR2022) read more