PointNet

Reported on 121 benchmarks across 14 tasks · 3 papers · 71 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Computer Vision68 results

Object DetectiononnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonScanObjectNN
Mean Accuracy· 2016-12-02
63.4
best: 93.8 (GPSFormer)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonScanObjectNN
Overall Accuracy· 2016-12-02
68.2
best: 97.2 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonIntrA
F1 score (5-fold)· 2016-12-02
0.684
best: 0.936 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40-C
Error Rate· 2016-12-02
0.283
best: 0.142 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 10-way (20-shot)
Overall Accuracy· 2016-12-02
35.2
best: 96.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 10-way (20-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 5-way (10-shot)
Overall Accuracy· 2016-12-02
51.97
best: 98 (PointGPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 5-way (10-shot)
Standard Deviation· 2016-12-02
12.1
best: 16 (PointNet++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 10-way (10-shot)
Overall Accuracy· 2016-12-02
46.6
best: 95 (Point-JEPA)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 10-way (10-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 5-way (20-shot)
Overall Accuracy· 2016-12-02
57.81
best: 99.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40 5-way (20-shot)
Standard Deviation· 2016-12-02
15.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonScanObjectNN
GFLOPs· 2016-12-02
0.5
best: 45 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonScanObjectNN
Number of params (M)· 2016-12-02
3.5
best: 34.2 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonScanObjectNN
Overall Accuracy (PB_T50_RS)· 2016-12-02
68
best: 92.64 (Mamba3D)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Semantic SegmentationonKITTI-360
mIoU Category· 2016-12-02
30.42
best: 74.08 (MinkowskiNet)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Semantic SegmentationonKITTI-360
miou· 2016-12-02
13.07
best: 58.3 (DeepViewAgg)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Object DetectiononnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonScanObjectNN
Mean Accuracy· 2016-12-02
63.4
best: 93.8 (GPSFormer)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonScanObjectNN
Overall Accuracy· 2016-12-02
68.2
best: 97.2 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonIntrA
F1 score (5-fold)· 2016-12-02
0.684
best: 0.936 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40-C
Error Rate· 2016-12-02
0.283
best: 0.142 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 10-way (20-shot)
Overall Accuracy· 2016-12-02
35.2
best: 96.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 10-way (20-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 5-way (10-shot)
Overall Accuracy· 2016-12-02
51.97
best: 98 (PointGPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 5-way (10-shot)
Standard Deviation· 2016-12-02
12.1
best: 16 (PointNet++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 10-way (10-shot)
Overall Accuracy· 2016-12-02
46.6
best: 95 (Point-JEPA)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 10-way (10-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 5-way (20-shot)
Overall Accuracy· 2016-12-02
57.81
best: 99.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40 5-way (20-shot)
Standard Deviation· 2016-12-02
15.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonScanObjectNN
GFLOPs· 2016-12-02
0.5
best: 45 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonScanObjectNN
Number of params (M)· 2016-12-02
3.5
best: 34.2 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonScanObjectNN
Overall Accuracy (PB_T50_RS)· 2016-12-02
68
best: 92.64 (Mamba3D)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Point Cloud SegmentationonPointCloud-C
mean Corruption Error (mCE)· 2016-12-02
1.178
best: 0.923 (GDANet)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononScanObjectNN
Mean Accuracy· 2016-12-02
63.4
best: 93.8 (GPSFormer)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononScanObjectNN
Overall Accuracy· 2016-12-02
68.2
best: 97.2 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononIntrA
F1 score (5-fold)· 2016-12-02
0.684
best: 0.936 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40-C
Error Rate· 2016-12-02
0.283
best: 0.142 (OmniVec2)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 10-way (20-shot)
Overall Accuracy· 2016-12-02
35.2
best: 96.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 10-way (20-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 5-way (10-shot)
Overall Accuracy· 2016-12-02
51.97
best: 98 (PointGPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 5-way (10-shot)
Standard Deviation· 2016-12-02
12.1
best: 16 (PointNet++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 10-way (10-shot)
Overall Accuracy· 2016-12-02
46.6
best: 95 (Point-JEPA)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 10-way (10-shot)
Standard Deviation· 2016-12-02
13.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 5-way (20-shot)
Overall Accuracy· 2016-12-02
57.81
best: 99.5 (ReCon++)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40 5-way (20-shot)
Standard Deviation· 2016-12-02
15.5
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononScanObjectNN
GFLOPs· 2016-12-02
0.5
best: 45 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononScanObjectNN
Number of params (M)· 2016-12-02
3.5
best: 34.2 (PCM)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononScanObjectNN
Overall Accuracy (PB_T50_RS)· 2016-12-02
68
best: 92.64 (Mamba3D)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40
Mean Accuracy· 2016-12-02
86
best: 92.4 (ULIP + PointMLP)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Shape Representation Of 3D Point CloudsonModelNet40
Overall Accuracy· 2016-12-02
89.2
best: 95.3 (PointGST)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40
Mean Accuracy· 2016-12-02
86
best: 92.4 (ULIP + PointMLP)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ClassificationonModelNet40
Overall Accuracy· 2016-12-02
89.2
best: 95.3 (PointGST)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40
Mean Accuracy· 2016-12-02
86
best: 92.4 (ULIP + PointMLP)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3D Point Cloud ReconstructiononModelNet40
Overall Accuracy· 2016-12-02
89.2
best: 95.3 (PointGST)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Object DetectiononnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
Object DetectiononnuScenes
mAOE
1.6
Object DetectiononnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
Object DetectiononnuScenes
mASE
0.58
best: 1 (qww)
Object DetectiononnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
Object DetectiononnuScenes
mAVE
2.21
3D Object DetectiononnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
3D Object DetectiononnuScenes
mAOE
1.6
3D Object DetectiononnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
3D Object DetectiononnuScenes
mASE
0.58
best: 1 (qww)
3D Object DetectiononnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
3D Object DetectiononnuScenes
mAVE
2.21

Methodology28 results

3DonnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
2D ClassificationonnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
2D Object DetectiononnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
16konnuScenes
mAAE· 2016-12-02
0.5
best: 1 (BirdNet+ (multisweep))
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
3DonnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
3DonnuScenes
mAOE
1.6
3DonnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
3DonnuScenes
mASE
0.58
best: 1 (qww)
3DonnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
3DonnuScenes
mAVE
2.21
2D ClassificationonnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
2D ClassificationonnuScenes
mAOE
1.6
2D ClassificationonnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
2D ClassificationonnuScenes
mASE
0.58
best: 1 (qww)
2D ClassificationonnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
2D ClassificationonnuScenes
mAVE
2.21
2D Object DetectiononnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
2D Object DetectiononnuScenes
mAOE
1.6
2D Object DetectiononnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
2D Object DetectiononnuScenes
mASE
0.58
best: 1 (qww)
2D Object DetectiononnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
2D Object DetectiononnuScenes
mAVE
2.21
16konnuScenes
NDS
0.16
best: 55.3 (LabelDistill)
16konnuScenes
mAOE
1.6
16konnuScenes
mAP
0.11
best: 45.1 (LabelDistill)
16konnuScenes
mASE
0.58
best: 1 (qww)
16konnuScenes
mATE
0.87
best: 1.06 (3D-GCK)
16konnuScenes
mAVE
2.21

Medical13 results

Semantic SegmentationonS3DIS Area5
mAcc· 2016-12-02
49
best: 81.6 (Sonata + PTv3)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonS3DIS
mAcc· 2016-12-02
66.2
best: 89.9 (Sonata + PTv3)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonKITTI-360
mIoU Category· 2016-12-02
30.42
best: 74.08 (MinkowskiNet)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonKITTI-360
miou· 2016-12-02
13.07
best: 58.3 (DeepViewAgg)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonIntrA
DSC (A)· 2016-12-02
49.59
best: 89.71 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonIntrA
DSC (V)· 2016-12-02
85
best: 97.29 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonIntrA
IoU (A)· 2016-12-02
37.75
best: 82.39 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonIntrA
IoU (V)· 2016-12-02
75.23
best: 94.82 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
Semantic SegmentationonS3DIS Area5
mIoU· 2020-12-16
41.1
best: 76 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
Semantic SegmentationonS3DIS
Mean IoU· 2020-12-16
47.6
best: 82.3 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
Semantic SegmentationonS3DIS
Mean IoU· 2020-12-16
47.6
best: 82.3 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
Semantic SegmentationonS3DIS
oAcc· 2019-04-16
78.5
best: 93.3 (Sonata + PTv3)
A-CNN: Annularly Convolutional Neural Networks on Point Clouds arXiv:1904.08017
Semantic SegmentationonShapeNet-Part
Instance Average IoU· 2016-12-02
83.7
best: 89.1 (GeomGCNN)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593

Audio13 results

10-shot image generationonS3DIS Area5
mAcc· 2016-12-02
49
best: 81.6 (Sonata + PTv3)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonS3DIS
mAcc· 2016-12-02
66.2
best: 89.9 (Sonata + PTv3)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonKITTI-360
mIoU Category· 2016-12-02
30.42
best: 74.08 (MinkowskiNet)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonKITTI-360
miou· 2016-12-02
13.07
best: 58.3 (DeepViewAgg)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonIntrA
DSC (A)· 2016-12-02
49.59
best: 89.71 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonIntrA
DSC (V)· 2016-12-02
85
best: 97.29 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonIntrA
IoU (A)· 2016-12-02
37.75
best: 82.39 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonIntrA
IoU (V)· 2016-12-02
75.23
best: 94.82 (3DMedPT)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593
10-shot image generationonS3DIS Area5
mIoU· 2020-12-16
41.1
best: 76 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
10-shot image generationonS3DIS
Mean IoU· 2020-12-16
47.6
best: 82.3 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
10-shot image generationonS3DIS
Mean IoU· 2020-12-16
47.6
best: 82.3 (Sonata + PTv3)
Point Transformer arXiv:2012.09164
10-shot image generationonS3DIS
oAcc· 2019-04-16
78.5
best: 93.3 (Sonata + PTv3)
A-CNN: Annularly Convolutional Neural Networks on Point Clouds arXiv:1904.08017
10-shot image generationonShapeNet-Part
Instance Average IoU· 2016-12-02
83.7
best: 89.1 (GeomGCNN)
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593

Graphs1 result

Point Cloud ClassificationonPointCloud-C
mean Corruption Error (mCE)· 2016-12-02
1.422
best: 0.455 (BeyondRPC)
SOTA
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation arXiv:1612.00593