I) ALL classes of MSR interleaved with CambridgeToyota II) Just CamToy classes + shortcuts III) Color-coding lookup chart, sequential IV) Hierarchy of CamToy classes - up for discussion ======================================================================== ======================================================================== I) // Below are the classes along with their colors - twice. // First, ALL the classes, with MSR interleaved with CamToyota, // where XXblah means we're not using that MSR class. // Second, same list, but only the CamToyota classes are listed, // and these include the short-cut keys we use ( "= " ). // I checked with the listing at the very bottom: these are sequential! =================================== Void 0 0 0 Building 128 0 0 XXgrass 0 128 0 Tree 128 128 0 XXcow 0 0 128 XXhorse 128 0 128 XXsheep 0 128 128 Sky 128 128 128 XXmountain 64 0 0 XXaeroplane 192 0 0 XXwater 64 128 0 XXface 192 128 0 Car 64 0 128 Archway 192 0 128 XXflower 64 128 128 SignSymbol 192 128 128 XXbird 0 64 0 XXbook 128 64 0 XXchair 0 192 0 XXmotorbike 128 192 0 XXperson 0 64 128 Road 128 64 128 XXcat 0 192 128 XXdog 128 192 128 Pedestrian 64 64 0 XXboat 192 64 0 Wall 64 192 0 VegetationMisc 192 192 0 Fence 64 64 128 Train 192 64 128 ParkingBlock 64 192 128 Column_Pole 192 192 128 TrafficCone 0 0 64 XXunused 128 0 64 Bridge 0 128 64 Misc_Text 128 128 64 Sidewalk 0 0 192 LaneMkgsDriv 128 0 192 Bicyclist 0 128 192 RoadShoulder 128 128 192 Tunnel 64 0 64 LaneMkgsNonDriv 192 0 64 Animal 64 128 64 Child 192 128 64 CartLuggagePram 64 0 192 MotorcycleScooter 192 0 192 SUVPickupTruck 64 128 192 Truck_Bus 192 128 192 TrafficLight 0 64 64 OtherMoving 128 64 64 ======================================================================== II) = d Void 0 0 0 = u Building 128 0 0 0 128 0 = e Tree 128 128 0 0 0 128 128 0 128 0 128 128 = k Sky 128 128 128 64 0 0 192 0 0 64 128 0 192 128 0 = c Car 64 0 128 Archway 192 0 128 64 128 128 = y SignSymbol 192 128 128 0 64 0 128 64 0 0 192 0 128 192 0 0 64 128 = r Road 128 64 128 0 192 128 128 192 128 = i Pedestrian 64 64 0 192 64 0 = w Wall 64 192 0 = v VegetationMisc 192 192 0 = f Fence 64 64 128 Train 192 64 128 = p ParkingBlock 64 192 128 Column_Pole 192 192 128 = t TrafficCone 0 0 64 128 0 64 = g Bridge 0 128 64 = m Misc_Text 128 128 64 = s Sidewalk 0 0 192 = l LaneMkgsDriv 128 0 192 = b Bicyclist 0 128 192 RoadShoulder 128 128 192 = n Tunnel 64 0 64 LaneMkgsNonDriv 192 0 64 = a Animal 64 128 64 Child 192 128 64 CartLuggagePram 64 0 192 MotorcycleScooter 192 0 192 SUVPickupTruck 64 128 192 Truck_Bus 192 128 192 TrafficLight 0 64 64 = o OtherMoving 128 64 64 ======================================================================== ======================================================================== III) // From ClassColorCoder's output (Jamie's) ========================================== 0 0 0 0 1 128 0 0 2 0 128 0 3 128 128 0 4 0 0 128 5 128 0 128 6 0 128 128 7 128 128 128 8 64 0 0 9 192 0 0 10 64 128 0 11 192 128 0 12 64 0 128 13 192 0 128 14 64 128 128 15 192 128 128 16 0 64 0 17 128 64 0 18 0 192 0 19 128 192 0 20 0 64 128 21 128 64 128 22 0 192 128 23 128 192 128 24 64 64 0 25 192 64 0 26 64 192 0 27 192 192 0 28 64 64 128 29 192 64 128 30 64 192 128 31 192 192 128 32 0 0 64 33 128 0 64 34 0 128 64 35 128 128 64 36 0 0 192 37 128 0 192 38 0 128 192 39 128 128 192 40 64 0 64 41 192 0 64 42 64 128 64 43 192 128 64 44 64 0 192 45 192 0 192 46 64 128 192 47 192 128 192 48 0 64 64 49 128 64 64 ======================================================================== ======================================================================== // IV) CamToyota Class hierarchy Note: so far, we've only used single entries from the list, so no nesting of categories (ie Road hasn't yet included RoadShoulder) Example: ----------(Semantic category)---------- A B C <-D Example explained: Pixels in the images are painted as A, B, and C, but could probably all go under by the name "A". There are 4 "Semantic categories" that don't appear as classes in the data, but if we had to describe the world with just 4 classes (or wanted 100% accuracy) these would be the 4 to get right. <-D = I think D is separate, but could go either way. ================================== ----------(Road)---------- Road 128 64 128 21 RoadShoulder 128 128 192 39 LaneMkgsDriv 128 0 192 37 LaneMkgsNonDriv 192 0 64 41 ----------(Fixed Objects)---------- Building 128 0 0 1 Wall 64 192 0 26 Bridge 0 128 64 34 <-Fence 64 64 128 28 Tree 128 128 0 3 VegetationMisc 192 192 0 27 Sidewalk 0 0 192 36 ParkingBlock 64 192 128 30 SignSymbol 192 128 128 15 Misc_Text 128 128 64 35 <-TrafficLight 0 64 64 48 Column_Pole 192 192 128 31 TrafficCone 0 0 64 32 ----------(Moving Objects)---------- Car 64 0 128 12 SUVPickupTruck 64 128 192 46 <-Truck_Bus 192 128 192 47 Train 192 64 128 29 Animal 64 128 64 42 Pedestrian 64 64 0 24 Child 192 128 64 43 CartLuggagePram 64 0 192 44 Bicyclist 0 128 192 38 MotorcycleScooter 192 0 192 45 OtherMoving 128 64 64 49 ----------(Ceiling)---------- Sky 128 128 128 7 Archway 192 0 128 13 Tunnel 64 0 64 40