Large-Scale 3D Shape Reconstruction and Segmentation from ShapeNet Core55
Computer Vision and Pattern Recognition
2017-10-31 v2
Abstract
We introduce a large-scale 3D shape understanding benchmark using data and annotation from ShapeNet 3D object database. The benchmark consists of two tasks: part-level segmentation of 3D shapes and 3D reconstruction from single view images. Ten teams have participated in the challenge and the best performing teams have outperformed state-of-the-art approaches on both tasks. A few novel deep learning architectures have been proposed on various 3D representations on both tasks. We report the techniques used by each team and the corresponding performances. In addition, we summarize the major discoveries from the reported results and possible trends for the future work in the field.
Cite
@article{arxiv.1710.06104,
title = {Large-Scale 3D Shape Reconstruction and Segmentation from ShapeNet Core55},
author = {Li Yi and Lin Shao and Manolis Savva and Haibin Huang and Yang Zhou and Qirui Wang and Benjamin Graham and Martin Engelcke and Roman Klokov and Victor Lempitsky and Yuan Gan and Pengyu Wang and Kun Liu and Fenggen Yu and Panpan Shui and Bingyang Hu and Yan Zhang and Yangyan Li and Rui Bu and Mingchao Sun and Wei Wu and Minki Jeong and Jaehoon Choi and Changick Kim and Angom Geetchandra and Narasimha Murthy and Bhargava Ramu and Bharadwaj Manda and M Ramanathan and Gautam Kumar and P Preetham and Siddharth Srivastava and Swati Bhugra and Brejesh Lall and Christian Haene and Shubham Tulsiani and Jitendra Malik and Jared Lafer and Ramsey Jones and Siyuan Li and Jie Lu and Shi Jin and Jingyi Yu and Qixing Huang and Evangelos Kalogerakis and Silvio Savarese and Pat Hanrahan and Thomas Funkhouser and Hao Su and Leonidas Guibas},
journal= {arXiv preprint arXiv:1710.06104},
year = {2017}
}