Official implementation of the MICCAI 2024
paper "TinyU-Net: Lighter yet Better U-Net with Cascaded Multi-Receptive Fields".
๐ฅ This paper has been invited for an ORAL
presentation in addition to a POSTER presentation.
- ISIC2018. The ISIC2018 dataset consists of images with skin disease lesions (2594 training images, 100 validation images, and 1000 test images).
- NCP. Lesion segmentation dataset of the CT slice images from the China Consortium of Chest CT Image Investigation (CC-CCII). A total of 750 CT slices from 150 COVID-19 patients were manually segmented into background, lung field, ground-glass opacity (GGO), and consolidation (CL).