English

Towards Fine-grained Large Object Segmentation 1st Place Solution to 3D AI Challenge 2020 -- Instance Segmentation Track

Computer Vision and Pattern Recognition 2024-04-05 v2

Abstract

This technical report introduces our solutions of Team 'FineGrainedSeg' for Instance Segmentation track in 3D AI Challenge 2020. In order to handle extremely large objects in 3D-FUTURE, we adopt PointRend as our basic framework, which outputs more fine-grained masks compared to HTC and SOLOv2. Our final submission is an ensemble of 5 PointRend models, which achieves the 1st place on both validation and test leaderboards. The code is available at https://github.com/zehuichen123/3DFuture_ins_seg.

Keywords

Cite

@article{arxiv.2009.04650,
  title  = {Towards Fine-grained Large Object Segmentation 1st Place Solution to 3D AI Challenge 2020 -- Instance Segmentation Track},
  author = {Zehui Chen and Qiaofei Li and Feng Zhao},
  journal= {arXiv preprint arXiv:2009.04650},
  year   = {2024}
}

Comments

Technical Report

R2 v1 2026-06-23T18:26:02.770Z