Model Zoo (Pretrained Models)
We provide the training logs & pretrained models in column our released trained with the improved training strategies proposed by PointNeXt. Existing public artifacts are linked through Google Drive. New release artifacts should also be mirrored on Hugging Face Hub with the layout, downloader, and checksum manifest described in Checkpoints and Hugging Face release layout.
Quick ModelNet40 checkpoint download after the HF mirror is published:
pip install pointnext_official
pointnext-download modelnet40-pointnext-s-c64 --output-dir ./hf_cache
TP: Throughput (instance per second) measured using an NVIDIA Tesla V100 32GB GPU and a 32 core Intel Xeon @ 2.80GHz CPU.
ScanObjectNN (Hardest variant) Classification
Throughput is measured with 128 x 1024 points.
| name | OA/mAcc (Original) | OA/mAcc (our released) | #params | FLOPs | Throughput (ins./sec.) |
|---|---|---|---|---|---|
| PointNet | 68.2 / 63.4 | 75.2 / 71.4 | 3.5M | 1.0G | 4212 |
| DGCNN | 78.1 / 73.6 | 86.1 / 84.3 | 1.8M | 4.8G | 402 |
| PointMLP | 85.4±1.3 / 83.9±1.5 | 87.7 / 86.4 | 13.2M | 31.4G | 191 |
| PointNet++ | 77.9 / 75.4 | 86.2 / 84.4 | 1.5M | 1.7G | 1872 |
| PointNeXt-S | 87.7±0.4 / 85.8±0.6 | 88.20 / 86.84 | 1.4M | 1.64G | 2040 |
| Pix4Point | 87.9 / 86.7 | 87.9 / 86.7 | 22.6M | 28.0G | - |
| PointVector | 87.8±0.4 / 86.2±0.5 | 88.17 / 86.69 | 1.55 | - | 901 |
S3IDS (6-fold) Segmentation
Throughput (TP) is measured with 16 x 15000 points.
| name | mIoU/OA/mAcc (Original) | mIoU/OA/mAcc (our released) | #params | FLOPs | TP |
|---|---|---|---|---|---|
| PointNet++ | 54.5 / 81.0 / 67.1 | 68.1 / 87.6 / 78.4 | 1.0M | 7.2G | 186 |
| PointNeXt-S | 68.0 / 87.4 / 77.3 | 68.0 / 87.4 / 77.3 | 0.8M | 3.6G | 227 |
| PointNeXt-B | 71.5 / 88.8 / 80.2 | 71.5 / 88.8 / 80.2 | 3.8M | 8.8G | 158 |
| PointNeXt-L | 73.9 / 89.8 / 82.2 | 73.9 / 89.8 / 82.2 | 7.1M | 15.2G | 115 |
| PointNeXt-XL | 74.9 / 90.3 / 83.0 | 74.9 / 90.3 / 83.0 | 41.6M | 84.8G | 46 |
| PointVector-L | 77.4 / 91.4 / 85.5 | 77.4 / 91.4 / 85.5 | 4.2M | 10.7G | 98 |
| PointVector-XL | 78.4 / 91.9 / 86.1 | 78.4 / 91.9 / 86.1 | 24.1M | 58.5G | 40 |
S3DIS (Area 5) Segmentation
Throughput (TP) is measured with 16 x 15000 points.
| name | mIoU/OA/mAcc (Original) | mIoU/OA/mAcc (our released) | #params | FLOPs | TP |
|---|---|---|---|---|---|
| PointNet++ | 53.5 / 83.0 / - | 63.6 / 88.3 / 70.2 | 1.0M | 7.2G | 186 |
| ASSANet | 63.0 / - /- | 65.8 / 88.9 / 72.2 | 2.4M | 2.5G | 228 |
| ASSANet-L | 66.8 / - / - | 68.0 / 89.7/ 74.3 | 115.6M | 36.2G | 81 |
| PointNeXt-S | 63.4±0.8 / 87.9±0.3 / 70.0±0.7 | 64.2 / 88.2 / 70.7 | 0.8M | 3.6G | 227 |
| PointNeXt-B | 67.3±0.2 / 89.4±0.1 / 73.7±0.6 | 67.5 / 89.4 / 73.9 | 3.8M | 8.8G | 158 |
| PointNeXt-L | 69.0±0.5 / 90.0±0.1 / 75.3±0.8 | 69.3 / 90.1 / 75.7 | 7.1M | 15.2G | 115 |
| PointNeXt-XL | 70.5±0.3 / 90.6±0.2 / 76.8±0.7 | 71.1 / 91.0 / 77.2 | 41.6M | 84.8G | 46 |
| Pix4Point | 69.6 / 89.9 / 75.2 | 69.6 / 89.9 / 75.2 | 23.7M | 190G | - |
| PointVector-L | 71.2 / 90.8 / 77.3 | 71.2 / 90.8 / 77.3 | - | - | - |
| PointVector-XL | 72.3 / 91.0 / 78.1 | 72.61 / 91.59 / 78.3 | 24.1M | 58.5G | 40 |
ShapeNetpart Part Segmentation
Throughput (TP) is measured with 64*2048 points.
| name | Ins. mIoU / Cat. mIoU (Original) | Ins. mIoU / Cat. mIoU (our released) |
|---|---|---|
| PointNet++ | 85.1/81.9 | |
| PointNeXt-S | 86.7±0.0 / 84.4±0.2 | 86.7 / 84.2 |
| PointNeXt-S (C=64) | 86.9±0.0 / 84.8±0.5 | 86.9 / 85.2 |
| PointNeXt-S (C=160) | 87.0±0.1 / 85.2±0.1 | 87.1 / 85.4 |
| Pix4Point | 86.8 / 85.6 | 86.8 / 85.6 |
| PointVector-S (C=64) | 86.9 / - | 86.9 / 85.05 |
ModelNet40 Classificaiton
| name | OA/mAcc (Original) | OA/mAcc (our released) | #params | FLOPs | Throughput (ins./sec.) |
|---|---|---|---|---|---|
| PointNet++ | 91.9 / - | 93.0 / 90.7 | 1.5M | 1.7G | 1872 |
| PointNeXt-S (C=64) | 93.7±0.3 / 90.9±0.5 | 94.0 / 91.1 | 4.5M | 6.5G | 2033 |
| PointVector-S (C=64) | 93.5±0.2 / 91.0±0.5 | 93.68 / 91.53 | - | - | - |