Haotian Zhang
Research Scientist at Apple
Former Research Intern at MSR AI
Ph.D. at University of Washington
Apple AI/ML
Biography
“Be Boundless.”
Haotian Zhang is a Research Scientist at Apple AI/ML, Visual Intelligence. His research aims to enable embodied agents to understand the outside world. To that end, he works on designing sensible modules that learn the effective representation of information from 2D/3D image data, as well as natural language. His recent work on GLIP&GLIPv2 has been accepted to the CVPR 2022 (Best Paper Finalist), and NeurIPS 2022. He also co-organized the ECCV 2022 workshop on Computer Vision in the Wild.
Prior to joining Apple, he obtained his Ph.D. in the Information Processing Lab at University of Washington, advised by Prof. Jenq-Neng Hwang, where he focused on monocular 3D object detection and multi-object tracking. He received his B.S. degree at Shanghai Jiao Tong University in 2017, supervised by Prof. Jun-Fa Mao.
He believes that living an interesting life is done by doing interesting things with interesting people, and that’s what he hopes to do 🔥.
Download CV here.
- Open-Vocabulary Object Detection
- Vision-and-Language Pre-training
- Large-scaled Pre-trained Models
- 3D Monocular Object Detection
- 2D/3D Multi-object Tracking
PhD in Elecricial & Computer Engineering, 2022
University of Washington
MSc in Applied Mathematics, 2021
University of Washington
MSc in Elecricial & Computer Engineering, 2019
University of Washington
BSc in Nano & Microelectronics, 2017
Shanghai Jiao Tong University
Recent News
[10/2022] Serving as session co-chair for ECCV CVinW Workshop and being responsible for ODinW. Full schedule here: https://computer-vision-in-the-wild.github.io/eccv-2022/.
[10/2022] Selected as one of the Young Scholar Award recipients for NeurIPS 2022.
[09/2022] One paper accepted by NeurIPS 2022: GLIPv2. A team effort to push CVinW
[08/2022] Updated GLIP Hugging Face Gradio Demo! Feel free to check it out!!!
[09/2022] Organizing ECCV Workshop Computer Vision in the Wild (CVinW), where two challenges Image Classification in the Wild (ICinW) and Object Detection in the Wild (ODinW) are hosted to evaluate the zero-shot, few-shot and full-shot performance of pre-trained vision models.
[03/2022] One paper accepted by CVPR 2022: GLIP as an Oral & Best Paper Finalist.
Featured Publications
Working Experience
