Flex-Convolution (Million-Scale Point-Cloud Learning Beyond Grid-Worlds)

2018/03/20 by Fabian Groh, Groh, Fabian, Patrick Wieschollek +3 · 3 citations
Computer Science · Engineering · #3D Shape Modeling and Analysis #Advanced Neural Network Applications #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Human Pose and Action Recognition

paper · pdf · doi:10.48550/arxiv.1803.07289

openalex publication_date 2018/03/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Traditional convolution layers are specifically designed to exploit the natural data representation of images -- a fixed and regular grid. However, unstructured data like 3D point clouds containing irregular neighborhoods constantly breaks the grid-based data assumption. Therefore applying best-practices and design choices from 2D-image learning methods towards processing point clouds are not readily possible. In this work, we introduce a natural generalization flex-convolution of the conventional convolution layer along with an efficient GPU implementation. We demonstrate competitive performance on rather small benchmark sets using fewer parameters and lower memory consumption and obtain significant improvements on a million-scale real-world dataset. Ours is the first which allows to efficiently process 7 million points concurrently.

Citations

• cuDNN: Efficient Primitives for Deep Learning
• Very Deep Convolutional Networks for Large-Scale Image Recognition
• ShapeNet: An Information-Rich 3D Model Repository
• A scalable active framework for region annotation in 3D shape collections
• Joint 2D-3D-Semantic Data for Indoor Scene Understanding
• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
• Detecting Functionality-Specific Vulnerabilities via Retrieving Individual Functionality-Equivalent APIs in Open-Source Repositories

Cited by

• Adaptive 3D Gaussian Splatting Video Streaming: Visual Saliency-Aware Tiling and Meta-Learning-Based Bitrate Adaptation
• Relation-Shape Convolutional Neural Network for Point Cloud Analysis
• SAMBLE: Shape-Specific Point Cloud Sampling for an Optimal Trade-Off Between Local Detail and Global Uniformity

Related

• PointCNN++: Performant Convolution on Native Points
• FPConv: Learning Local Flattening for Point Convolution
• Optimizing Sparse Convolution on GPUs with CUDA for 3D Point Cloud Processing in Embedded Systems
• KPConv: Flexible and Deformable Convolution for Point Clouds
• TorchSparse: Efficient Point Cloud Inference Engine