Research

Continual federated active learning for 3D detection + LiDAR-camera calibration.

Graduate Researcher · Oct 2024 – Jan 2026

Riverside, CA

Benchmarked FCOS3D in continual federated active env (Flower.ai + MMDetection3D, uniform/entropy/confidence/margin/coreset/CALD sampling)
Collected multi-agent CARLA interactions for multi-agent learning
Rebuilt Command Input Imitation Learning pipeline with custom CARLA image + position data in PyTorch
Created 150+ static/dynamic camera/LiDAR snapshots for targetless calibration
Benchmarked CalibNet on KITTI, NuScenes, CalibDB as baseline for BevCalib

Publication

Weiduo Yuan, Jerry Li, Justin Yue, Divyank Shah, Konstantinos Karydis, Hang Qiu

9th Conference on Robot Learning (CoRL 2025), Seoul, Korea

Publication

Bo Wu, Ruoshen Mo, Justin Yue, Yanyu Zhang, Janice Nguyen, Guoyuan Wu, Amit Roy-Chowdhury, Matthew J. Barth, Hang Qiu

European Conference on Computer Vision (ECCV 2026)

Divyank Shah — acknowledged in the paper's acknowledgements for effort and participation in data collection and annotation.

RESEARCHRESEARCH