3D Point Cloud Classification On Modelnet40
Metrics
Overall Accuracy
Results
Performance results of various models on this benchmark
Comparison Table
| Model Name | Overall Accuracy |
|---|---|
| point-convolutional-neural-networks-by | 92.3 |
| pct-point-cloud-transformer | 93.2 |
| let-images-give-you-more-point-cloud-cross | 94.4 |
| masked-autoencoders-for-point-cloud-self | 94.0 |
| point-cloud-classification-using-content | 93.5 |
| interpolated-convolutional-networks-for-3d | 93.0 |
| point2vec-for-self-supervised-representation | 94.8 |
| relation-shape-convolutional-neural-network | 92.9 |
| pointnet-deep-hierarchical-feature-learning | 90.7 |
| dspoint-dual-scale-point-cloud-recognition | 93.5 |
| geometric-feedback-network-for-point-cloud | 93.8 |
| sagemix-saliency-guided-mixup-for-point | 90.3 |
| point-m2ae-multi-scale-masked-autoencoders | 94.0 |
| parameter-is-not-all-you-need-starting-from | 93.8 |
| attention-based-point-cloud-edge-sampling | 93.5 |
| Model 16 | 94.9 |
| pointnext-revisiting-pointnet-with-improved | 94.0 |
| perceiver-general-perception-with-iterative | - |
| learning-geometry-disentangled-representation | 93.8 |
| generative-and-discriminative-voxel-modeling | - |
| dynamic-local-geometry-capture-in-3d | 92.1 |
| mvtn-multi-view-transformation-network-for-3d | 93.8 |
| fg-net-fast-large-scale-lidar-point | 93.8 |
| point-cloud-pre-training-by-mixing-and | 93.39 |
| kpconv-flexible-and-deformable-convolution | 92.9 |
| anisotropic-separable-set-abstraction-for | 92.9 |
| rethinking-masked-representation-learning-for | 94.5 |
| pointcutmix-regularization-strategy-for-point | 93.4 |
| advanced-feature-learning-on-point-clouds | 93.3 |
| 3d-point-cloud-classification-and | 91.6 |
| paconv-position-adaptive-convolution-with | 93.9 |
| point-transformer-v2-grouped-vector-attention | 94.2 |
| pointmixer-mlp-mixer-for-point-cloud | 93.6 |
| general-purpose-deep-point-cloud-feature | 91.7 |
| deltaconv-anisotropic-point-cloud-learning | 93.8 |
| shellnet-efficient-point-cloud-convolutional | 93.1 |
| pointnet-deep-learning-on-point-sets-for-3d | 89.2 |
| point-cloud-pre-training-by-mixing-and | 93.31 |
| positional-prompt-tuning-for-efficient-3d | 93.88 |
| multi-view-convolutional-neural-networks-for | 90.1 |
| revisiting-point-cloud-classification-with-a | 93.9 |
| learning-inner-group-relations-on-point | 94.1 |
| lcpformer-towards-effective-3d-point-cloud | 93.6 |
| mamba3d-enhancing-local-features-for-3d-point | 95.1 |
| point-transformer | 92.8 |
| escape-from-cells-deep-kd-networks-for-the | 90.6 |
| adacrossnet-adaptive-dynamic-loss-weighting | 93.1 |
| generative-and-discriminative-voxel-modeling | - |
| volumetric-and-multi-view-cnns-for-object | 89.2 |
| point-voxel-transformer-an-efficient-approach | 94.0 |
| 3d-medical-point-transformer-introducing | 93.4 |
| pcp-mae-learning-to-predict-centers-for-point | 94.2 |
| pointcnn-convolution-on-x-transformed-points | 92.2 |
| rethinking-the-compositionality-of-point | 94.5 |
| rethinking-network-design-and-local-geometry-1 | 94.5 |
| pointconv-deep-convolutional-networks-on-3d | 92.5 |
| attention-based-point-cloud-edge-sampling | 93.8 |
| 3d-shapenets-a-deep-representation-for | - |
| p2p-tuning-pre-trained-image-models-for-point | 94.0 |
| point-transformer-1 | 93.7 |
| so-net-self-organizing-network-for-point | 90.9 |
| dynamic-graph-cnn-for-learning-on-point | 92.9 |
| a-deep-dive-into-explainable-self-supervised | 94.2 |
| pointasnl-robust-point-clouds-processing | 93.2 |
| point-m2ae-multi-scale-masked-autoencoders | 92.9 |
| surface-representation-for-point-clouds | 94.7 |
| 190408017 | 92.6 |
| modelnet-o-a-large-scale-synthetic-dataset | 94.0 |
| revisiting-point-cloud-shape-classification | 93.9 |
| parameter-efficient-fine-tuning-in-spectral | 95.3 |
| point2sequence-learning-the-shape | 92.6 |
| rethinking-masked-representation-learning-for | 94.3 |
| Model 73 | 93.1 |
| pointscnet-point-cloud-structure-and | 93.7 |
| contrast-with-reconstruct-contrastive-3d | 94.7 |
| point-jepa-a-joint-embedding-predictive | 94.1±0.1 |
| ulip-learning-unified-representation-of | 93.4 |
| regularization-strategy-for-point-cloud-via | 93.5 |
| ulip-learning-unified-representation-of | 94.7 |
| decoupled-local-aggregation-for-point-cloud | 94.0 |
| towards-compact-3d-representations-via-point | 94.5 |
| sagemix-saliency-guided-mixup-for-point | 93.3 |
| spatio-temporal-self-supervised | 93.1 |
| shapellm-universal-3d-object-understanding | 95.0 |
| polynet-polynomial-neural-network-for-3d | 92.42 |
| pointgrid-a-deep-network-for-3d-shape | 92.0 |
| ckconv-learning-feature-voxelization-for | 94.0 |
| instance-aware-dynamic-prompt-tuning-for-pre | 94.4 |
| deepgcns-making-gcns-go-as-deep-as-cnns | 93.6 |
| escape-from-cells-deep-kd-networks-for-the | 91.8 |
| dense-resolution-network-for-point-cloud | 93.1 |
| local-spectral-graph-convolution-for-point | 92.1 |
| spidercnn-deep-learning-on-point-sets-with | 92.4 |
| person-re-identification-in-the-3d-space | 93.3 |
| implicit-autoencoder-for-point-cloud-self | 94.2 |
| regress-before-construct-regress-autoencoder | 94.1 |
| manifold-net-using-manifold-learning-for | 93.0 |
| crossmoco-multi-modal-momentum-contrastive | 91.49 |
| point-bert-pre-training-3d-point-cloud | 93.8 |
| point-is-a-vector-a-feature-representation-in | 93.5 |
| points-to-patches-enabling-the-use-of-self | 92.6 |
| point-jepa-a-joint-embedding-predictive | 93.8±0.2 |
| dynamic-edge-conditioned-filters-in | 87.4 |
| pra-net-point-relation-aware-network-for-3d | 93.7 |
| 3d-jepa-a-joint-embedding-predictive | 94.0 |
| point-planenet-plane-kernel-based | 92.1 |
| walk-in-the-cloud-learning-curves-for-point | 94.2 |
| sagemix-saliency-guided-mixup-for-point | 93.6 |
| dualmlp-a-two-stream-fusion-model-for-3d | 93.7 |
| ulip-learning-unified-representation-of | 94.1 |