Semantic Segmentation on SIFT Flow (test)
64Mean AccuracyRegion-based end-to-end
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Region-based end-to-endYear=20162016.07 | 64 | 84.3 | — | — | — | |
| SVM+JCSupervision=Fully supervised, Backbone=VGG16, Calibration Method=joint calibration (JC)2015.07 | 59.2 | — | — | — | — | |
| SVCNet2019.09 | 58.2 | 89.1 | 46.3 | — | — | |
| DAG-RNN+CRF2019.09 | 57.8 | 87.8 | 44.8 | — | — | |
| EigenYear=20152016.07 | 55.7 | 86.8 | — | — | — | |
| CaesarYear=20152016.07 | 55.6 | — | — | — | — | |
| SVM+JCSupervision=Fully supervised, Backbone=AlexNet, Calibration Method=joint calibration (JC)2015.07 | 55.6 | — | — | — | — | |
| FCN-8s2016.05 | 53.9 | 85.9 | 41.2 | 77.2 | — | |
| FCN-8s2019.09 | 53.9 | 85.9 | 41.2 | — | — | |
| Contextual Deep CRFs2015.04 | 53.4 | 88.1 | — | — | 44.9 | |
| Piecewise2019.09 | 53.4 | 88.1 | 44.9 | — | — | |
| Sharma [12]Year=20152016.07 | 52.8 | 80.9 | — | — | — | |
| Sharma et al.Supervision=Fully supervised, citation=[29]2015.07 | 52.8 | — | — | — | — | |
| FCN-16sarchitecture=two-headed version2014.11 | 51.7 | 85.2 | 39.5 | 76.1 | — | |
| FCN-16s2015.04 | 51.7 | 85.2 | — | — | 39.5 | |
| LongYear=20152016.07 | 51.7 | 85.2 | — | — | — | |
| Long et al.Supervision=Fully supervised2015.07 | 51.7 | — | — | — | — | |
| Farabet et al. [8]variant=1, sampling=class-balanced samples2014.11 | 50.8 | 72.3 | — | — | — | |
| Farabet et al. naturalScale=multi-scale, Training distribution=natural frequency2016.05 | 50.8 | 72.3 | — | — | — | |
| Farabet et al.sampling=balance2015.04 | 50.8 | 72.3 | — | — | — | |
| FarabetYear=20132016.07 | 50.8 | 78.5 | — | — | — | |
| Farabet et al.Supervision=Fully supervised2015.07 | 50.8 | — | — | — | — | |
| GeorgeYear=20152016.07 | 50.1 | 81.7 | — | — | — | |
| George et al.Supervision=Fully supervised2015.07 | 50.1 | — | — | — | — | |
| YangYear=20142016.07 | 48.7 | 79.8 | — | — | — | |
| Yang et al.Supervision=Fully supervised2015.07 | 48.7 | — | — | — | — | |
| Yang et al.2019.09 | 48.7 | 79.8 | — | — | — | |
| Sharma [11]Year=20142016.07 | 48 | 79.6 | — | — | — | |
| Sharma et al.Supervision=Fully supervised, citation=[28]2015.07 | 48 | — | — | — | — | |
| KekeçYear=20142016.07 | 45.8 | 70.4 | — | — | — | |
| Kekeç et al.Supervision=Fully supervised2015.07 | 45.8 | — | — | — | — | |
| SVM+JCSupervision=Weakly supervised, Backbone=VGG16, Calibration Method=joint calibration (JC)2015.07 | 44.8 | — | — | — | — | |
| Xu et al.Supervision=Weakly supervised, Transductive=true, citation=[42]2015.07 | 41.4 | — | — | — | — | |
| Tighe et al. [34]variant=1, model=exemplar SVM2014.11 | 41.1 | 75.6 | — | — | — | |
| Tighe et al. SVM2016.05 | 41.1 | 75.6 | — | — | — | |
| Tighe et al.2015.04 | 41.1 | 75.6 | — | — | — | |
| Tighe [14]Year=20132016.07 | 41.1 | 78.6 | — | — | — | |
| Tighe et al.Supervision=Fully supervised, citation=[35]2015.07 | 41.1 | — | — | — | — | |
| Tighe et al.2019.09 | 41.1 | 75.6 | — | — | — | |
| ShuaiYear=20152016.07 | 39.7 | 80.1 | — | — | — | |
| Shuai et al.Supervision=Fully supervised2015.07 | 39.7 | — | — | — | — | |
| Tighe et al. [34]variant=2, model=SVM + MRF2014.11 | 39.2 | 78.6 | — | — | — | |
| Tighe et al. SVM+MRF2016.05 | 39.2 | 78.6 | — | — | — | |
| Tighe et al.model=MRF2015.04 | 39.2 | 78.6 | — | — | — | |
| SVM+JCSupervision=Weakly supervised, Backbone=AlexNet, Calibration Method=joint calibration (JC)2015.07 | 37.4 | — | — | — | — | |
| Xu et al.Supervision=Weakly supervised, Transductive=false, citation=[42]2015.07 | 35 | — | — | — | — | |
| SinghYear=20132016.07 | 33.8 | 79.2 | — | — | — | |
| Sharma et al.2019.09 | 33.6 | 79.6 | — | — | — | |
| Zhang et al.Supervision=Weakly supervised, citation=[45]2015.07 | 32.3 | — | — | — | — | |
| GattaYear=20142016.07 | 32.1 | 78.7 | — | — | — | |
| PinheiroYear=20142016.07 | 30 | 76.5 | — | — | — | |
| Pinheiro et al.Supervision=Fully supervised2015.07 | 30 | — | — | — | — | |
| Pinheiro et al.model=RCNN3 (o3)2014.11 | 29.8 | 77.7 | — | — | — | |
| Pinheiro et al.Scale=multi-scale, Recurrence=RCNN32016.05 | 29.8 | 77.7 | — | — | — | |
| Pinheiro et al.2015.04 | 29.8 | 77.7 | — | — | — | |
| Pinheiro et al.2019.09 | 29.8 | 77.7 | — | — | — | |
| Farabet et al. [8]variant=2, sampling=natural frequency samples2014.11 | 29.6 | 78.5 | — | — | — | |
| Farabet et al. balancedScale=multi-scale, Training distribution=class-balanced2016.05 | 29.6 | 78.5 | — | — | — | |
| Farabet et al.2015.04 | 29.6 | 78.5 | — | — | — | |
| Farabet et al.2019.09 | 29.6 | 78.5 | — | — | — | |
| Tighe [13]Year=20102016.07 | 29.4 | 76.9 | — | — | — | |
| Tighe et al.Supervision=Fully supervised, citation=[33]2015.07 | 29.1 | — | — | — | — | |
| SVMSupervision=Fully supervised, Backbone=AlexNet, Calibration Method=uncalibrated SVM scores2015.07 | 28.7 | — | — | — | — | |
| Xu et al.Supervision=Weakly supervised, citation=[41]2015.07 | 27.9 | — | — | — | — | |
| SVM+PSSupervision=Fully supervised, Backbone=AlexNet, Calibration Method=Platt scaling (PS)2015.07 | 27.7 | — | — | — | — | |
| Zhang et al.Supervision=Weakly supervised, citation=[44]2015.07 | 27.7 | — | — | — | — | |
| GouldYear=20142016.07 | 25.7 | 78.4 | — | — | — | |
| ByeonYear=20152016.07 | 22.6 | 68.7 | — | — | — | |
| Byeon et al.Supervision=Fully supervised2015.07 | 22.6 | — | — | — | — | |
| SVMSupervision=Weakly supervised, Backbone=AlexNet, Calibration Method=uncalibrated SVM scores2015.07 | 21.2 | — | — | — | — | |
| Vezhnevets et al.Supervision=Weakly supervised, citation=[40]2015.07 | 21 | — | — | — | — | |
| SVM+PSSupervision=Weakly supervised, Backbone=AlexNet, Calibration Method=Platt scaling (PS)2015.07 | 16.8 | — | — | — | — | |
| Vezhnevets et al.Supervision=Weakly supervised, citation=[39]2015.07 | 14 | — | — | — | — | |
| Liu et al.2014.11 | — | 76.7 | — | — | — | |
| Liu et al.2016.05 | — | 76.7 | — | — | — | |
| Liu et al.2015.04 | — | 76.7 | — | — | — | |
| Liu et al.2019.09 | — | 76.7 | — | — | — |