Lily Xianling Zhang

Senior Machine Learning Research Scientist, Technical Project Lead. Ford Autonomy, 3D Perception and Prediction

prof_pic.jpg

Ford Greenfield Labs, building b

3251 Hillview Ave

Palo Alto, CA 94304

Lily (Xianling) Zhang is a Machine Learning Research Scientist at Ford Autonomy, Palo Alto, CA, located at Ford Greenfield Labs, Stanford Research Park. She is working on Perception and Prediction for autonomous driving tasks.

My research interests include

  • 3D perception and motion prediction
  • Geometric scene understanding based data manipulation(scene relighting) to significantly improve the performance of SOTA CenterTrack algorithm.
  • Dataset bias identification using perceptual similarity based data clustering for large scale dataset meta labeling, similar data searching, balanced data selection for training set curation.
  • Data augmentation via GANs that achieves consistent improvement over multple AV task: parking slot detection, highway lane segmentation, and monodepth estimation.

news

Apr 26, 2023 At Computer Vision Summit 2023 , I am going to give a keynote on Deep learning-based data-centric solutions for vision tasks. Link to the talk.
Mar 21, 2023 At NVIDIA GTC 2023, I am going to give a talk using AI based data algorithms to achieve autonomous driving dataset diversity. We will have live discussion on Tuesday, Mar 21. 2:00 PM - 2:50 PM PDT. Link to the talk (S51407)
Feb 22, 2023 At AI Accelerator Institute 2023 , I am going to give a talk on Data misconceptions, challenges and solutions

selected publications

  1. lanerf.png
    LANe: Lighting-Aware Neural Fields for Compositional Scene Synthesis
    Akshay Krishnan, Amit Raj, Xianling Zhang, and 5 more authors
    2023
  2. simbar.gif
    SIMBAR: Single Image-Based Scene Relighting for Effective Data Augmentation for Automated Driving Vision Tasks
    Xianling Zhang, Nathan Tseng, Ameerah Syed, and 2 more authors
    In CVPR, Jun 2022
  3. deeppic.png
    deepPIC: Deep Perceptual Image Clustering for Identifying Bias in Vision Datasets
    Nikita Jaipuria, Katherine Stevo, Xianling Zhang, and 4 more authors
    In CVPR, Jun 2022
  4. synthetic_aug.gif
    Deflating Dataset Bias Using Synthetic Data Augmentation
    Nikita Jaipuria, Xianling Zhang, Rohan Bhasin, and 5 more authors
    In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, Jun 2020

service

I regularly serve as a reviewer for CVPR, NeurIPS, ICCV.