×
This website is optimized for a screen size of 1025x800px or higher. If you are accessing this website with a lower resolution, some features are not fully functional.

Result Visualization

New: WildDash 2 with 4256 public frames, new labels & panoptic GT!  
See also: RailSem19 dataset for rail scene understanding.

Excerpt from submitted algorithm results using the standard cityscapes color coding (white is used for unknown labels; e.g. ScanNet). The order in the list corresponds to the leaderboard order.

Left side: Three positive test cases
  • wd0010; An busy inner-city scene, some overexposure
  • wd0026; City scene with near-miss collision
  • wd0123; Night scene with heavy rain
Right side: Three negative test cases
  • wd0141; RGB/BGR channels swapped
  • wd0146; Pure image transmission noise
  • wd0152; Aquarium scene; out-of-scope scene
go back to leaderboard
RUSH_ROB
NL_ROI_ROB
MaskRCNN-R-50-FPN-GN
MRCNN_CS
MRCNN_VSCMLab_ROB
MaskRCNN_ROB
MRCNN++_VSCMLab_ROB
Sem2Ins
MRCNN_K
BAMRCNN_ROB
Cached Oct. 24, 2025, 6:12 a.m. UTC+0