SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation

Benchmark Model Rank Results
crowd-counting-on-ucf-qnrfEncoder-Decoder#19MAE: 270
lesion-segmentation-on-anatomical-tracings-ofSegNet#3IoU: 0.1911Precision: 0.3938Recall: 0.2532Dice: 0.2767
lesion-segmentation-on-university-of-waterlooSegNet#2Dice score: 0.854 ±0.088
real-time-semantic-segmentation-on-camvidSegNet#27mIoU: 46.4%Frame (fps): 4.6Time (ms): 217
semantic-segmentation-on-ade20kSegNet#221Validation mIoU: 21.64
semantic-segmentation-on-camvidSegNet#17Mean IoU: 46.4%
semantic-segmentation-on-cityscapesSegNet#94Mean IoU (class): 57.0%
semantic-segmentation-on-skyscapes-dense-1SegNet#5Mean IoU: 23.14
semantic-segmentation-on-tlcgisSegNet#4IoU: 77.80
thermal-image-segmentation-on-mfn-datasetSegNet#43mIOU: 42.3