no code implementations • 1 Apr 2024 • Hao Chen, Yuqi Hou, Chenyuan Qu, Irene Testini, Xiaohan Hong, Jianbo Jiao
While many existing datasets focus on scene understanding from a certain perspective (e. g. egocentric or third-person views), our dataset offers a panoptic perspective (i. e. multiple viewpoints with multiple data modalities).
1 code implementation • ICCV 2023 • Hao Chen, Chenyuan Qu, Yu Zhang, Chen Chen, Jianbo Jiao
It is understandable as the model is designed to learn paired mapping (e. g. from a noisy image to its clean version).
Ranked #1 on Denoising on CBSD68 sigm75