This is the unofficial code of Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes. which achieve state-of-the-art trade-off between accuracy and speed on cityscapes and camvid, without using inference acceleration and extra data
I ran eval.py following this project's manual.: $ python tools/eval.py --cfg experiments/cityscapes/ddrnet23_slim.yaml
I got this result.: MeanIU: 0.7783, Pixel_Acc: 0.9601, Mean_Acc: 0.8548, Class IoU: [0.98051424 0.84613866 0.92252055 0.51403614 0.61611903 0.64109395 0.7016854 0.77541595 0.92519328 0.66229634 0.9446796 0.81535788 0.63319689 0.9494065 0.79650392 0.88227362 0.81294668 0.60278734 0.76526594] Mins: 1
I would like to know the meaning of the above matrix.