sparse-convolution

Minkowski Engine is an auto-diff neural network library for high-dimensional sparse tensors

Caffe for Sparse and Low-rank Deep Neural Networks

[ICLR'23 Spotlight🔥] The first successful BERT/MAE-style pretraining on any convolutional network; Pytorch impl. of "Designing BERT for Convolutional Networks: Sparse and Hierarchical Masked Modeling"

Focal Sparse Convolutional Networks for 3D Object Detection (CVPR 2022, Oral)

ISBNet: a 3D Point Cloud Instance Segmentation Network with Instance-aware Sampling and Box-aware Dynamic Convolution (CVPR 2023)

Unofficial PyTorch implementation of the paper: "CenterNet3D: An Anchor free Object Detector for Autonomous Driving"

[NeurIPS 2022, T-PAMI 2023] Efficient Spatially Sparse Inference for Conditional GANs and Diffusion Models

[IROS 2020] Indoor Scene Recognition in 3D

[ECCV 2024] 3D Small Object Detection with Dynamic Spatial Pruning

[CVPR24] MaGGIe: Mask Guided Gradual Human Instance Matting

[WACV'25] Official implementation of "PK-YOLO: Pretrained Knowledge Guided YOLO for Brain Tumor Detection in Multiplane MRI Slices".

Neural Medial Axis Approximation of Point Clouds for 3D Tree Skeletonization

Lidar Bridge Classification

Open Source Project for 3D Semantic Segmentation

A fast MATLAB toolbox for N-dimensional sparse arrays.

spconv-Triton is a fast sparse convolution library with full operator support (SubM conv, Conv3D, Transposed Conv, Pooling). It is hardware-agnostic and runs on GPUs supporting Triton.

Fork of NVIDIA MinkowskiEngine modernized for CUDA 12.8+/Blackwell (RTX 50-series), NumPy 2.0, PyTorch 2.x prebuilt wheels, fp16/bf16 autocast, fused gather/scatter. Best-effort, no support commitment.

Sparse ConvLSTM for Point Cloud Semantic Segmentation

Efficient convolution for sparse data on FPGAs

An implementation of Sparse Layers in TensorFlow 2. x.