's Research Page!
  • Home
  • Publications
  • Activities
Picture

​ Xinge Zhu 

PhD
MMLab, CUHK

Google Scholar / Github / CV

Email:  zhuxinge123 at gmail.com

New Events

  • 2023.02: NEW!!Two papers are accepted by CVPR2023.
  • 2023.01: NEW!!Two papers are accepted by ICRA2023.
  • 2022.11: One paper is accepted by AAAI2023.
  • 2022.08: Vision-Centric BEV Perception: A Survey is released and project page at this URL.
  • 2022.07: One paper is accepted by ECCV2022.
  • 2022.06: A simple multi-view baseline, MV-FCOS3D++, obtained the runner-up in the Waymo Camera-Only 3D Detection Challenge
  • 2022.05: PVKD (1st place in the challenge of SemanticKITTI LiDAR Semantic Segmentation Challenge) is accepted by CVPR2022. Code has been released
  • 2022.04: STCrowd (A Multimodal Dataset for Crowded Pedestrian Perception) is accepted by CVPR2022.
  • 2022.03: TransFusion (1st place in the challenge of nuScenes tracking task) is accepted by CVPR2022. Code has been released
  • 2022.03: Three papers are accepted by CVPR2022.
  • 2021.11: Extended version of AdaStereo is accepted by IJCV
  • 2021.09: FCOS3D got the best paper award on the ICCV 3DODI workshop! Code has been released
  • 2021.09: FCOS3D is accepted by 3DODI workshop, ICCV. SIDE (Stereo 3D Detection) is accepted by WACV2022. PGD is accepted by CoRL2021;
  • 2021.07: Extended version of Cylinder3D is accepted by TPAMI
  • 2021.03: Three papers are accepted by CVPR2021 (1 Oral and 2 Poster). Cylinder3D is accepted as Oral presentation
  • 2020.12: Our team achieves the 2nd place in the challenge of nuScenes LiDAR segmentation with Cylinder3D, and the best camera-only method in nuScenes 3D detection with PGD
  • 2020.11: We release the code of Cylinder3D, which achieves the 1st place in the leaderboard of SemanticKITTI Semantic segmentation. Based on Cylinder3D, we also achieve the 1st place in the leaderboard of SemanticKITTI Panoptic segmentation.

  • About Me

    Greetings! My name is Xinge Zhu. I obtained my PhD degree at Multimedia lab in CUHK. My supervisor is Prof. Dahua Lin.

    My research lies at computer vision and deep learning, particularly in Scene Understanding and Perception, Point-cloud 3D Detection and Segmentation, etc.

    During my PhD, I spent wonderful time at SenseTime, Baidu and Inceptio. I also work closely with Prof. Yuexin Ma, Prof. Ruigang Yang, Dr. Jianping Shi and Prof. Dinesh Manocha.


    Selected Awards and Honors

  • nuScenes Tracking, 1st place (2022.03)
  • nuScenes 3D Detection, Best camera-only method (2020.12)
  • nuScenes LiDAR Segmentation, 2nd place (2020.12)
  • SemanticKITTI Semantic Segmentation, Rank 1st (2020.11)
  • SemanticKITTI Panoptic Segmentation, Rank 1st
  • Lyft 3D Detection Challenge, Golden Medal (rank 4 among 500+ teams, 2019.7)
  • nuScenes 3D Detection Challenge, Runner-up (rank 2 among 13 teams, 2019.3)

  • Talks

  • Talk at SUSTech about 2D and 3D Detection, Mar 2022.
  • Talk at ShanghaiTech University about 2D and 3D Detection, Spring 2021.
  • Zhidx Open Talk about Large-scale Point-cloud Segmentation , Summer 2021.
  • Open Talk at JiangTech for CVPR 2021, Summer 2021.