AI & Visual Computing Laboratory (AI Lab)

Lab Lead

A/Prof. Cuong Pham — Lead Principal Investigator & Associate Professor

Associate Professor Cuong Pham is the Dean of Computer Science & Engineering faculty (PTIT) and a Research Scientist at Qualcomm AI Research (acquiring  VinAI’s Research division since April 2025).  Previously, he was the Dean of PTIT’s Artificial Intelligence faculty, a Research Scientist at VinAI Research (from November 2019 to March 2025); a Marie Curie Research Fellow at Philips Research, High-Tech campus Eindhoven, the Netherlands; a Research Associate at Open Lab, Newcastle University, UK; a Research Assistant at Jornada Experimental Range, the United State Department of Agriculture, USA; and a Teaching Assistant at PTIT. I have regularly published papers in top-tier AI conferences and journals such as CVPR, NeurIPS, ICCV, ECCV, ICML, TPAMI, AAAI. My research interests include Deep Generative models, Computer Vision, Human Activity Recognition, Deep Learning and Pervasive Healthcare. 

Researchers & Collaborators

  • Postdoctoral Researchers & PhD Scholars: Focusing on scalable multi-expert systems, diffusion model optimization, 3D vision, and generative video synthesis.
  • Research Alumni: Including research residents and doctoral students placed at top international institutions and research labs (e.g., University of Wisconsin-Madison, VinAI Research).

Research Areas

  • Deep Generative Models & Efficient AI
    Research on accelerated single-step/few-step diffusion generation, model inversion, controllable image and video editing, post-training quantization, and model compression for real-time deployment.
  • Computer Vision & Scene Understanding
    Open-vocabulary 3D instance segmentation, unsupervised image/video deblurring with dense-matching references, and quality-controllable image manipulation.
  • Human Activity Recognition & Exemplar Analysis
    Exemplar identification, few-shot human action counting in complex unconstrained environments, and multi-sensor activity modeling.
  • Pervasive Healthcare & Sensor Systems
    IoT-enabled automated respiratory symptom monitoring, ambient sensing, unobtrusive wearable tracking (smart shoes), and real-time fall detection using heterogeneous sensor fusion.
    • Graph & Network Intelligence
      Active learning frameworks for incomplete networks and reinforced multi-expert algorithms for complex network influence maximization.

    Selected Publications

    • NeurIPS 2026 — “ControlJEPA: Principled Trajectory Regularization via Lyapunov Tube Loss”
    • NeurIPS 2026 — “FlashControl: One-Step Controllable Generator via Distillation Friendly Single-Stream Teachers”
    • CVPR 2026 — “InverFill: One-Step Inversion for Enhanced Few-Step Diffusion Inpainting”
    • CVPR 2026 — “BluRef: Unsupervised Image Deblurring with Dense-Matching References”
    • CVPR 2025 — “SwiftEdit: Lightning Fast Text-guided Image Editing via One-step Diffusion”
    • CVPR 2025 — “Any3DIS: Class-Agnostic 3D Instance Segmentation by 2D Mask Tracking”
    • AAAI 2025 — “REM: A Scalable Reinforced Multi-Expert Framework for Multiplex Influence Maximization”
    • ECCV 2024 — “SwiftBrush v2: Make Your One Step Diffusion Better Than Its Teacher”
    • UAI 2024 — “Active Learning Framework for Incomplete Networks”
    • CVPR 2024 — “Blur2Blur: Blur Conversion for Unsupervised Image Deblurring on Unknown Domains”
    • CVPR 2024 — “Open3dis: Open-vocabulary 3D Instance Segmentation with 2D Mask Guidance”
    • AAAI 2024 — “Count What You Want: Exemplar Identification and Few-Shot Counting of Human Actions in the Wild”
    • CVPR 2023 — “HyperCUT: Video Sequence from a Single Blurry Image using Unsupervised Ordering”
    • NeurIPS 2022 — “QC-StyleGAN: Quality Controllable Image Generation and Manipulation”