FreDSNet: Joint Monocular Depth and Semantic Segmentation with Fast Fourier Convolutions

Benchmark Model Rank Results
depth-estimation-on-stanford2d3d-panoramicFreDSNet#1RMSE: 0.2727absolute relative error: 0.0952
semantic-segmentation-on-stanford2d3d-1FreDSNet#17mIoU: 46.1%mAcc: 63.1