Researcher Wei Zhan

Current Projects

Stimulate Reasoning Abilities of LLM in Interactive Prediction and Planning
Tracking Anyone and Anycar for 3D Auto-Labeling from Video Inputs
A Unified Generative Framework for 3D Driving Scenarios
Integrated Fine-Grained 3D Reconstruction and Narrow-Space Planning for Autonomous Parking
Truckload Optimization via Task Planning based on Deep Reinforcement Learning

Past Projects

3D Object Detection Enhanced by Temporal Multi-View Input
Efficient and Consistent Joint Multi-Agent Prediction for Interactive Planning
Online Automatic Construction and Update of HD Maps with Rich Semantics
Computationally Efficient Option-based Monte Carlo Planning for POMDP
Multi-Modal Self-Supervised Pre-training for Label-Efficient 3D Perception
Automatic Inference of Lane Topology and Connecting Geometry for Scene Understanding, High-Definition Map Construction and Representation
Data Efficiency for Point Cloud Perception Model
Embracing Uncertainties in 3D Object Detection -- from Annotation and Benchmarking to Failure Analysis and Data Augmentation
Automatic Semantics Extraction and Representation for High-Definition Map Construction and Scene Understanding
Efficient Model for Large-Scale Point Cloud Perception