Xin Liu

PhD Candidate @TU Delft Computer Vision Lab.

prof_pic.jpg

I’m a final-year PhD candidate at the Computer Vision Lab of Delft University of Technology, supervised by Dr. Jan van Gemert.

My research interests include efficient deep learning for video understanding, e.g. video object detection/tracking, segmentation, and action recognition; optimization for efficient training of DNNs on large-volume temporal data. Because I’m interested in video data learning and efficient deep learning models, I’m also exploring possibilities of efficient NeRF, NeRF for Video and Video Diffusion Models.

In my free time, I enjoy learning new things, such as coffee making and pottery, and playing tennis and board games.

If you’re interested to collaborate, feel free to drop me an email.

I’m on the job market for full-time positions. Feel free to contact me. :)

news

Jul 18, 2023 Our paper “Objects do not disappear: Video object detection by single-frame object location anticipation” got accepted to ICCV 2023.
Mar 3, 2023 Invited talk at Microsoft ASG UK, “Efficient Video Object Detection by Forecasting Object Locations”.
Dec 6, 2022 I’m reviewing for Journal of Visual Communication and Image Representation.
Jun 30, 2022 I’m attending ICVSS 2022 summer school.
May 11, 2022 I’m reviewing for ECCV 2022.
Dec 16, 2021 I’m reviewing for CVPR 2022.
Mar 31, 2021 I’ll be joining Tencent YouTu Lab as a Research Intern, working on unsupervised anomaly object detection.
Mar 12, 2021 Our paper “No frame left behind: Full Video Action Recognition” got accepted to CVPR 2021.

selected publications

  1. ICCV 2023
    Objects do not disappear: Video object detection by single-frame object location anticipation
    Xin Liu, Fatemeh Karimi Nejadasl, Jan C. Gemert, and 2 more authors
    2023
  2. ICCV-W 2023
    Video BagNet: short temporal receptive fields increase robustness in long-term action recognition
    Ombretta Strafforello, Xin Liu, Klamer Schutte, and 1 more author
    2023
  3. CVPR 2021
    No frame left behind: Full video action recognition
    Xin Liu, Silvia L Pintea, Fatemeh Karimi Nejadasl, and 2 more authors
    2021
  4. ICPR 2020
    WeightAlign: Normalizing Activations by Weight Alignment
    Xiangwei* Shi, Yunqiang* Li, Xin* Liu, and 1 more author
    2021
  5. ICCV-W 2019
    Cross domain image matching in presence of outliers
    Xin Liu, Seyran Khademi, and Jan Van Gemert
    2019