Kaicheng Yu

Tenure Track Assistant Professor, Autonomous Intelligence Lab, Westlake University

prof_pic.jpg

Autonomous Intelligence Lab

Westlake University

Hangzhou, China

We are hiring at all levels!

Current openings:

  • PhD Student, starting from Fall 2025.
  • 3-6 months algorithm engineer internship towards realistic challenges in autonomous driving scenarios
  • PostDoc

Please shoot an email with your resume to me if you are interested!

Short Bio

Dr. Kaicheng Yu is an Assistant Professor at Westlake University, where he founded the Autonomous Intelligence Lab (AutoLab) in 2023. Prior to his academic appointment, he conducted research at Intel Labs and Alibaba DAMO Academy. He was awarded the Qualcomm Innovation Fellowship Europe in 2019. Dr. Yu completed his Bachelor’s degree with First Class Honour from the University of Hong Kong in 2016 and earned his Ph.D. from École Polytechnique Fédérale de Lausanne (EPFL) in 2021. His research has been published in leading computer vision and machine learning venues, including CVPR, ICCV, ECCV, ICML, ICLR, and NeurIPS, as well as prominent journals such as IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). His research focuses on developing next-generation AI systems for autonomous intelligence, with applications on autonomous driving and AI for Science. Additionally, Dr. Yu serves as the Co-founder and Chief Scientist of KMind, a company developing agent-enhanced internet solutions to improve user efficiency through personalized local search.

His ultimate research goal is to build a system that incorporates automatic machine learning to evolve with a human-in-the-loop feedback to achieve generic 3D perception across multi-modalities for robotic vision systems. Specifically, his works include:

  • Towards robust 3D perception & action in autonomous driving scenarios in Bird’s Eye View fashion and cross-modality 3D generation;
  • Cross modalities of 3D data generation in the wild, including camera and LiDAR sensors;
  • Automatically improves the performance of neural architectures, including neural architecture search, active learning, and lifelong learning.

Academic Services

  • YOCSEF Hangzhou Academic Chair
  • Conference Tutorial: ACM MM 2024
  • Conference reviewer: CVPR, ICCV, ECCV, BMVC, ICML, ICLR, NeurIPS, MM, AAAI
  • Journal reviewer: IJCV, TPAMI, TIP, ACM-TMM, TMLR

news

Nov 6, 2024 I am invited to give a talk at Australia National University, hosted by Prof. Miaomiao Liu, titled ‘Large Visual-Centric Models in Society Autolab Perspective’.
Nov 1, 2024 I organized a tutorial at ACM Multimedia, Large Vision-Language Model in Society.
Oct 20, 2024 We have one paper accepted to NeurIPS 2024!
Jun 18, 2024 We have two papers accepted to CVPR 2024!
Aug 1, 2023 I officially left Alibaba and joined Westlake University as a tenure tracked assistant professor.
Jun 18, 2023 I am invited to give a talk at UBC, titled ‘Reducing human efforts in visual perception’.
Apr 23, 2023 I am invited to give a talk on China3DV about our recent progress on robust 3D detection
Mar 28, 2023 My personal web page is online, this site will be heavily adapted for a while!
Aug 7, 2021 I obtained my PhD diploma from EPFL.

selected publications

  1. Statistically-motivated Second-order Pooling
    Kaicheng Yu, and Mathieu Salzmann
    In European Conference on Computer Vision, 2018
  2. An Analysis of Super-Net Heuristics in Weight-Sharing NAS
    Kaicheng Yu, Rene Ranftl, and Mathieu Salzmann
    IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023
  3. Evaluating the Search Phase of Neural Architecture Search
    Kaicheng Yu, Christian Sciuto, Martin Jaggi, Claudiu Musat, and Mathieu Salzmann
    In International Conference on Learning Representations, Nov 2020
  4. Landmark Regularization: Ranking Guided Super-Net Training in Neural Architecture Search
    Kaicheng Yu, Rene Ranftl, and Mathieu Salzmann
    In IEEE Conference of Computer Vision and Pattern Recognition, Apr 2021
  5. BEVFusion: A Simple and Robust LiDAR-Camera Fusion Framework
    Tingting Liang, Hongwei Xie, Kaicheng Yu, Zhongyu Xia, Zhiwei Lin, Yongtao Wang, Tao Tang, Bing Wang, and Zhi Tang
    In Annual Conference on Neural Information Processing Systems, Nov 2022
  6. Benchmarking the Robustness of LiDAR-Camera Fusion for 3D Object Detection
    Kaicheng Yu, Tang Tao, Hongwei Xie, Zhiwei Lin, Zhongwei Wu, Zhongyu Xia, Tingting Liang, Haiyang Sun, Jiong Deng, Dayang Hao, Yongtao Wang, Xiaodan Liang, and Bing Wang
    In IEEE Conference on Computer Vision and Pattern Recognition Workshops, May 2023