Fast Sparse 3D Convolution Network with VDB

원문 초록

We proposed a new Convolution Neural Network implementation optimized for sparse 3D data inference. This implementation uses NanoVDB as the data structure to store the sparse tensor. It leaves a relatively small memory footprint while maintaining high performance. We demonstrate that this architecture is around 20 times faster than the state-of-the-art dense CNN model on a high-resolution 3D object classification network.

함께 살펴볼 논문

공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.

arXiv:2410.10084공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2024-10-14

PointNet with KAN versus PointNet with MLP for 3D Classification and Segmentation of Point Sets

arXiv:2408.08753공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2024-08-16

PCP-MAE: Learning to Predict Centers for Point Masked Autoencoders

arXiv:2408.06693공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2024-08-13

DC3DO: Diffusion Classifier for 3D Objects

arXiv:2405.01413공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2024-05-02

MiniGPT-3D: Efficiently Aligning 3D Point Clouds with Large Language Models using 2D Priors

arXiv:2403.05950공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2024-03-09

Classifying Objects in 3D Point Clouds Using Recurrent Neural Network: A GRU LSTM Hybrid Approach

arXiv:2312.07039공통 태그 3D Object Classification공식 코드 · Papers with Code 등록 2023-12-12

Open-Pose 3D Zero-Shot Learning: Benchmark and Challenges

토론

이 논문과 연결된 포럼 글이 아직 없습니다.

포럼에서 토론 찾기