Thang Doan Email & Phone Number

Los Angeles, Ca, Us

Los Angeles, Ca

Los Angeles, California, United States

Multimodal Search & Retrievalo Lead and Develop Computer Vision initiatives for Composed Image Retrieval o Auto-Eval pipeline design for error mode analysiso I-140 Approved for EB-1A Green Card

San Francisco Bay Area

-- Vision-Language Model / Prompt-tuning o Multi-Modality for Few-Shot Incremental Learning o UP-DP: Unsupervised Prompt Learning for Data Pre-Selection with Vision-Language Models (NeurIPS 2023)-- Open-World Problems o Hyp-OW: Exploiting Hierarchical Structure Learning with Hyperbolic Distance Enhances Open World Object Detection (AAAI 2024) o Deployed Segmentation model (CLIP-S4) on real self-driving data-- OOD Detection o GradOrth: A Simple yet Efficient Out-of-Distribution Detection with Orthogonal Projection of Gradients (NeurIPS 2023)o Awarded O1 visa

Montreal, Canada Area

-- Lifelong Learning o Continual Learning Beyond a Single Model (CoLLAs 2023) o A Theoretical Analysis of Catastrophic Forgetting through the NTK Overlap Matrix (AISTATS 2021) o Organized first workshop on Theory of Continual Learning (ICML 2021) -- Reinforcement Learning o Building a Subspace of Policies for Scalable Continual Learning" (ICLR 2023) o Low-Rank Representation of Reinforcement Learning Policies (JAIR 2022) o Regularized Inverse Reinforcement Learning (ICLR 2021) o Deep Reinforcement and InfoMax Learning (Neurips 2020) o Organized first worshop on Self-SUpervised Learning for RL (ICLR 2021)

Doctor Of Philosophy - Phd, Operation Management (Om)

-- Reinforcement Learning o Leveraging exploration in off-policy algorithms via normalizing flows (CoRL 2019) -- GANs o Multi-objective.

Msc, Mathematical Modelling

French Engineering Degree, Mechanical Engineering

Classe Preparatoire Aux Grandes Ecoles (Pcsi - Psi ), Natural Sciences

Mathematics, Physics, Chemistry, Industrial Science