Skip to content
View ahxmeds's full-sized avatar
🍄
🍄
23 followers · 35 following

Achievements

Achievement: YOLOAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Achievements

Achievement: YOLOAchievement: Pull Sharkx2Achievement: Arctic Code Vault Contributor

Block or report ahxmeds

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
ahxmeds/README.md

Hello, I'm Shadab 👋

I am a Research Scientist at Huawei Canada (2026-Present), where I work on diffusion- and flow-matching-based models for real-time, interactive video generation on mobile devices. My broader research background spans generative modeling for images, inverse problems, medical anomaly detection, and efficient deep learning.

Email LinkedIn GitHub

Current Focus

  • Interactive video generation using diffusion and flow matching models.
  • User-guided and controllable generative AI systems for visual content.

Prior Research Background

  • Postdoctoral Researcher, UBC (2025-2026): developed a theoretically grounded generative framework that mitigates ill-conditioned and uneven learning dynamics in flow-based generative models, enabling faster convergence and higher-quality image generation.
  • PhD Physics, UBC (2020-2025): developed deep learning and generative AI methods for solving inverse problems and image reconstruction, segmentation, synthesis, and enhancement, with applications in medical imaging and scientific computing.

Skills

Generative AI & Video

Diffusion Models Flow Matching Video Diffusion Wan2.1 WanAnimate SCAIL-2 Knowledge Distillation LoRA Variational Autoencoders

Machine Learning & Medical Imaging

PyTorch Hugging Face Diffusers Transformers MONAI PyTorch Geometric scikit-learn OpenCV SimpleITK

Training, Acceleration & Infrastructure

CUDA FlashAttention Distributed Training DDP FSDP Linux Docker Slurm Git Weights & Biases Conda SSH and HPC PyTorch Lightning Quantization ONNX Runtime ExecuTorch TensorRT Mobile AI Deployment

Programming & Scientific Computing

Python C++ MATLAB Mathematica NumPy Pandas Jupyter LaTeX

GitHub Snapshot

GitHub stats Top languages by repo Top languages by commit

Pinned Loading

  1. multiscale-gradient-estimation multiscale-gradient-estimation Public

    Official implementation of Multiscale Gradient Estimation (MGE) and Full-Multiscale training for reduced compute training of CNNs via multiscale gradient estimation (TMLR 2026).

    Python 6 2

  2. DAWN-FM DAWN-FM Public

    Data-aware and noise-informed flow matching for solving inverse problems

    Python 2

  3. IgCONDA-PET IgCONDA-PET Public

    Weakly-supervised PET anomaly detection using implicitly-guided attention-conditional counterfactual diffusion modeling.

    Python 3

  4. magnetic-inversion-dictionary magnetic-inversion-dictionary Public

    Unrolled dictionary learning for recovering 3D distribution of magnetic susceptibility from Earth's magnetic subsurface data.

    Jupyter Notebook 1 2

  5. microsoft/lymphoma-segmentation-dnn microsoft/lymphoma-segmentation-dnn Public

    Python 25 5

  6. thyroidiomics thyroidiomics Public

    Classification of thyroid diseases from scintigraphy images using a two-step method: thyroid segmentation and disease classification using the extracted radiomics features from the segmented regions.

    Python 7 2