Prompt Guided Transformer for Multi-Task Dense Prediction

Benchmark Model Rank Results
boundary-detection-on-nyu-depth-v2PGT (Swin-S)#2odsF: 78.04
boundary-detection-on-nyu-depth-v2PGT (Swin-T)#3odsF: 77.05
semantic-segmentation-on-nyu-depth-v2PGT (Swin-S)#65Mean IoU: 46.43
semantic-segmentation-on-nyu-depth-v2PGT (Swin-T)#76Mean IoU: 41.61
monocular-depth-estimation-on-nyu-depth-v2PGT (Swin-S)#72RMSE: 0.5468
monocular-depth-estimation-on-nyu-depth-v2PGT (Swin-T)#77RMSE: 0.59