Results for PTv3 - PPT

Submission info

Full name Point Transformer V3 with Point Prompt Training
Publication title Towards Large-scale 3D Representation Learning with Multi-dataset Point Prompt Training
Publication authors Xiaoyang Wu, Zhuotao Tian, Xin Wen, Bohao Peng, Xihui Liu, Kaicheng Yu, Hengshuang Zhao
Publication venue CVPR 2024
Publication url https://arxiv.org/abs/2308.09718
Project link https://github.com/Pointcept/Pointcept
Hardware A100 x 8
Code Code available with license: MIT
Use other data Yes: ScanNet200 (ScanNet), Structured3D
Last edited 22 May, 2024

Method details

3D semantic segmentation results

mIoU wall ceiling floor table door ceiling lamp cabinet blinds curtain chair storage cabinet office chair bookshelf whiteboard window box window frame monitor shelf doorframe pipe heater kitchen cabinet sofa windowsill bed shower wall trash can book plant blanket tv computer tower kitchen counter refrigerator jacket electrical duct sink bag picture pillow towel suitcase backpack crate keyboard rack toilet paper printer poster painting microwave board shoes socket bottle bucket cushion basket shoe rack telephone file folder cloth blind rail laptop plant pot exhaust fan cup coat hanger light switch speaker table lamp air vent clothes hanger kettle smoke detector container power strip slippers paper bag mouse cutting board toilet paper paper towel pot clock pan tap jar soap dispenser binder bowl tissue box whiteboard eraser toilet brush spray bottle headphones stapler marker
TOP-1 0.46451 0.83044 0.90892 0.92661 0.77533 0.70245 0.92110 0.36684 0.87677 0.78120 0.71171 0.54180 0.84112 0.49474 0.81025 0.63062 0.50799 0.42085 0.87533 0.25418 0.39625 0.72929 0.80163 0.55844 0.81965 0.68930 0.81283 0.21244 0.77903 0.17173 0.90555 0.74512 0.98351 0.55681 0.21827 0.83470 0.69195 0.31614 0.77993 0.42759 0.51495 0.39421 0.61718 0.58865 0.59171 0.14592 0.82923 0.00522 0.91950 0.29280 0.44793 0.14854 0.08275 0.89645 0.00067 0.40644 0.22786 0.38202 0.32748 0.02061 0.00742 0.30908 0.78144 0.00243 0.00000 0.62921 0.75378 0.53628 0.74222 0.43985 0.08396 0.46623 0.07785 0.84698 0.03484 0.05611 0.41436 0.42351 0.00233 0.37251 0.35401 0.15918 0.74961 0.16331 0.33505 0.26318 0.00000 0.73257 0.30207 0.44711 0.04335 0.73120 0.00000 0.11802 0.24738 0.39027 0.67825 0.05622 0.19248 0.05873 0.00000
TOP-3 0.71097 0.96941 0.99132 0.98364 0.91599 0.95699 0.96908 0.63808 0.98793 0.98369 0.88254 0.80808 0.97272 0.68476 0.97110 0.91796 0.70699 0.88670 0.91225 0.37772 0.96917 0.97902 0.91777 0.84469 0.93779 0.89604 0.98503 0.83792 0.94224 0.68692 0.97219 0.98142 0.99153 0.75191 0.60936 0.97091 0.80371 0.94600 0.96270 0.73061 0.75600 0.71202 0.81120 0.76728 0.86332 0.34370 0.93671 0.19009 0.99558 0.61817 0.58159 0.78351 0.49273 0.96365 0.12604 0.87613 0.69684 0.56051 0.77031 0.64175 0.16239 0.96246 0.93958 0.23119 0.01851 0.94522 0.96701 0.87616 0.94907 0.56468 0.31096 0.90238 0.47051 0.96671 0.30690 0.07551 0.76573 0.73679 0.11095 0.72066 0.71588 0.51721 0.96372 0.57781 0.44970 0.31174 0.00424 0.78500 0.63412 0.77371 0.21140 0.90784 0.00000 0.34817 0.58597 0.66423 0.84219 0.28026 0.73360 0.10045 0.98525