Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Datasets/ScanNet

ScanNet

ImagesRGB-DCustomIntroduced 2017-01-01

ScanNet is an instance-level indoor RGB-D dataset that includes both 2D and 3D data. It is a collection of labeled voxels rather than points or objects. Up to now, ScanNet v2, the newest version of ScanNet, has collected 1513 annotated scans with an approximate 90% surface coverage. In the semantic segmentation task, this dataset is marked in 20 classes of annotated 3D voxelized objects.

Source: A Review of Point Cloud Semantic Segmentation Image Source: http://www.scan-net.org/

Benchmarks

10-shot image generation/val mIoU 10-shot image generation/test mIoU 10-shot image generation/PQ 10-shot image generation/PQ_th 10-shot image generation/PQ_st 10-shot image generation/Average Accuracy 10-shot image generation/3DIoU 2D Semantic Segmentation/Average Recall 2D Semantic Segmentation/mIoU 3D/RMSE 3D/absolute relative error 3D/3DIoU 3D/Chamfer Distance 3D/L1 3D Character Animation From A Single Photo/Average Recall 3D Instance Segmentation/mAP 3D Reconstruction/3DIoU 3D Reconstruction/Chamfer Distance 3D Reconstruction/L1 Animation/Average Recall Continual Semantic Segmentation/mIoU Depth Estimation/RMSE Depth Estimation/absolute relative error Instance Segmentation/mAP Panoptic Segmentation/PQ Panoptic Segmentation/PQ_th Panoptic Segmentation/PQ_st Scene Parsing/Average Recall Scene Segmentation/Average Accuracy Scene Segmentation/3DIoU Semantic Segmentation/val mIoU Semantic Segmentation/test mIoU Semantic Segmentation/PQ Semantic Segmentation/PQ_th Semantic Segmentation/PQ_st Semantic Segmentation/Average Accuracy Semantic Segmentation/3DIoU Zero-Shot Learning/HmIoU

Related Benchmarks

ScanNet(v2)/3D Instance Segmentation/mAP ScanNet(v2)/3D Instance Segmentation/mAP @ 50 ScanNet(v2)/3D Instance Segmentation/mAP@25 ScanNet(v2)/3D Instance Segmentation/mRec ScanNet(v2)/Instance Segmentation/mAP ScanNet(v2)/Instance Segmentation/mAP @ 50 ScanNet(v2)/Instance Segmentation/mAP@25 ScanNet(v2)/Instance Segmentation/mRec ScanNet++/10-shot image generation/Top-1 IoU ScanNet++/10-shot image generation/Top-3 IoU ScanNet++/16k/mAP@0.25 ScanNet++/16k/mAP@0.5 ScanNet++/2D Classification/mAP@0.25 ScanNet++/2D Classification/mAP@0.5 ScanNet++/2D Object Detection/mAP@0.25 ScanNet++/2D Object Detection/mAP@0.5 ScanNet++/3D/mAP@0.25 ScanNet++/3D/mAP@0.5 ScanNet++/3D Instance Segmentation/mAP ScanNet++/3D Object Detection/mAP@0.25 ScanNet++/3D Object Detection/mAP@0.5 ScanNet++/3D Semantic Segmentation/Top-1 IoU ScanNet++/3D Semantic Segmentation/Top-3 IoU ScanNet++/Instance Segmentation/mAP ScanNet++/Novel View Synthesis/LPIPS ScanNet++/Novel View Synthesis/PSNR ScanNet++/Novel View Synthesis/SSIM ScanNet++/Object Detection/mAP@0.25 ScanNet++/Object Detection/mAP@0.5 ScanNet++/Semantic Segmentation/Top-1 IoU ScanNet++/Semantic Segmentation/Top-3 IoU ScanNet++ (trained on 3DMatch)/3D Point Cloud Interpolation/Recall ( correspondence RMSE below 0.2)ScanNet++ (trained on 3DMatch)/Point Cloud Registration/Recall ( correspondence RMSE below 0.2)ScanNet200/10-shot image generation/test mIoU ScanNet200/10-shot image generation/val mIoU ScanNet200/3D Instance Segmentation/mAP ScanNet200/3D Instance Segmentation/mAP@25 ScanNet200/3D Instance Segmentation/mAP@50 ScanNet200/3D Open-Vocabulary Instance Segmentation/AP Common ScanNet200/3D Open-Vocabulary Instance Segmentation/AP Head ScanNet200/3D Open-Vocabulary Instance Segmentation/AP Tail ScanNet200/3D Open-Vocabulary Instance Segmentation/AP25 ScanNet200/3D Open-Vocabulary Instance Segmentation/AP50 ScanNet200/3D Open-Vocabulary Instance Segmentation/mAP ScanNet200/3D Semantic Segmentation/test mIoU ScanNet200/3D Semantic Segmentation/val mIoU ScanNet200/Instance Segmentation/mAP ScanNet200/Instance Segmentation/mAP@25 ScanNet200/Instance Segmentation/mAP@50 ScanNet200/Semantic Segmentation/test mIoU ScanNet200/Semantic Segmentation/val mIoU ScanNetV1/Instance Segmentation/mAP@0.25 ScanNetV2/10-shot image generation/Mean IoU ScanNetV2/10-shot image generation/Mean IoU (test)ScanNetV2/10-shot image generation/Mean IoU (val)ScanNetV2/10-shot image generation/PQ ScanNetV2/10-shot image generation/Params (M)ScanNetV2/10-shot image generation/Pixel Accuracy ScanNetV2/10-shot image generation/RQ ScanNetV2/10-shot image generation/SQ ScanNetV2/10-shot image generation/mIoU ScanNetV2/16k/mAP@0.25 ScanNetV2/16k/mAP@0.5 ScanNetV2/2D Classification/mAP@0.25 ScanNetV2/2D Classification/mAP@0.5 ScanNetV2/2D Object Detection/mAP@0.25 ScanNetV2/2D Object Detection/mAP@0.5 ScanNetV2/2D Panoptic Segmentation/PQ ScanNetV2/3D/Delta < 1.25 ScanNetV2/3D/absolute relative error ScanNetV2/3D/mAP@0.25 ScanNetV2/3D/mAP@0.5 ScanNetV2/3D Instance Segmentation/NoC@80 ScanNetV2/3D Object Detection/mAP@0.25 ScanNetV2/3D Object Detection/mAP@0.5 ScanNetV2/3D Semantic Segmentation/mIoU ScanNetV2/Depth Estimation/Delta < 1.25 ScanNetV2/Depth Estimation/absolute relative error ScanNetV2/Instance Segmentation/NoC@80 ScanNetV2/Instance Segmentation/mAP@0.50 ScanNetV2/Interactive 3D Instance Segmentation/NoC@80 ScanNetV2/Object Detection/mAP@0.25 ScanNetV2/Object Detection/mAP@0.5 ScanNetV2/Panoptic Segmentation/PQ ScanNetV2/Panoptic Segmentation/Params (M)ScanNetV2/Panoptic Segmentation/RQ ScanNetV2/Panoptic Segmentation/SQ ScanNetV2/Semantic Segmentation/Mean IoU ScanNetV2/Semantic Segmentation/Mean IoU (test)ScanNetV2/Semantic Segmentation/Mean IoU (val)ScanNetV2/Semantic Segmentation/PQ ScanNetV2/Semantic Segmentation/Params (M)ScanNetV2/Semantic Segmentation/Pixel Accuracy ScanNetV2/Semantic Segmentation/RQ ScanNetV2/Semantic Segmentation/SQ ScanNetV2/Semantic Segmentation/mIoU ScanNetV2/Surface Normals Estimation/% < 11.25 ScanNetV2/Surface Normals Estimation/% < 22.5 ScanNetV2/Surface Normals Estimation/% < 30 ScanNetV2/Surface Normals Estimation/Mean Angle Error

Statistics

Papers: 1,595
Benchmarks: 38

Links

Tasks

10-shot image generation 2D Panoptic Segmentation 2D Semantic Segmentation 3D 3D Character Animation From A Single Photo 3D Instance Segmentation 3D Object Detection 3D Reconstruction 3D Semantic Instance Segmentation Animation Continual Semantic Segmentation Depth Estimation Generalized Zero-Shot Learning Instance Segmentation Interactive 3D Instance Segmentation Interactive 3D Instance Segmentation -Trained on Scannet40 - Evaluated on Scannet40 Panoptic Segmentation Scene Parsing Scene Recognition Scene Segmentation Semantic Segmentation Surface Normals Estimation Unsupervised 3D Semantic Segmentation Zero-Shot Learning Zero-shot 3D Point Cloud Classification