Writing

Notes & Deep Dives

A growing collection of short technical write-ups and long-form project deep dives — spanning vision-language models, federated learning, and applied medical imaging.

3 deep dives 0 notes
  1. Deep dive ICCV 2025

    PRISM: Debiasing Vision-Language Models via LLM-Guided Embedding Projection

    A data-free, task-agnostic debiasing framework for CLIP-style vision-language models.
  2. Deep dive AAAI 2025

    FedGaLA: Federated Unsupervised Domain Generalization via Gradient Alignment

    First framework for unsupervised federated domain generalization, grounded in a gradient-alignment theory.
  3. Deep dive Multimedia Tools and Applications

    COVID-CXNet: Open Chest X-ray Dataset and Detection Model for COVID-19

    One of the earliest large-scale open datasets and detection models for COVID-19 on frontal chest X-rays.