The Scene type classification task involves classifying a scan into 13 scene types.

Evaluation and metrics

Our evaluation ranks all methods according to recall (TP/(TP+FN)) as well as the PASCAL VOC intersection-over-union metric (IoU = TP/(TP+FP+FN)), where TP, FP, and FN are the numbers of true positive, false positive, and false negative predictions, respectively.



This table lists the benchmark results for the scene type classification scenario.




Method Infoavg recallapartmentbathroombedroom / hotelbookstore / libraryconference roomcopy/mail roomhallwaykitchenlaundry roomliving room / loungemiscofficestorage / basement / garage
sort bysort bysort bysort bysort bysort bysort bysorted bysort bysort bysort bysort bysort bysort by
3DASPP-SCE0.691 20.500 10.938 20.824 31.000 11.000 10.500 21.000 10.857 20.500 20.556 30.000 20.812 20.500 1
multi-taskpermissive0.700 10.500 11.000 10.882 20.500 21.000 11.000 10.500 21.000 11.000 10.778 10.000 20.938 10.000 2
Shengyu Huang, Mikhail Usvyatsov, Konrad Schindler: Indoor Scene Recognition in 3D. IROS 2020
SE-ResNeXt-SSMA0.498 30.000 40.812 30.941 10.500 20.500 30.500 20.500 20.429 40.500 20.667 20.500 10.625 30.000 2
Abhinav Valada, Rohit Mohan, Wolfram Burgard: Self-Supervised Model Adaptation for Multimodal Semantic Segmentation. arXiv
resnet50_scannet0.353 40.250 30.812 30.529 40.500 20.500 30.000 40.500 20.571 30.000 40.556 30.000 20.375 40.000 2