Utkarsh DoshiMasked Autoencoders Enable Efficient Knowledge Distillers: Paper ExplainedDistilling pre-trained model? Then this article is for you.Mar 9Mar 9
Utkarsh DoshiI-JEPA: ExplainedJEPA is a new architecture for self-supervised learning by Yann LeCun intended to overcome key limitations of the most advanced AI systems.Mar 41Mar 41
Utkarsh DoshiUnsupervised Segmentation? Yes.Ever thought of doing unsupervised segmentation? If yes, then this is the right article.Feb 231Feb 231
Utkarsh DoshiSegment Anything Model (SAM): ExplainedExplanation with emphasis on architecture and and a bit on both the process of building the dataset and zero-shot capabilities.Dec 25, 20233Dec 25, 20233
Utkarsh DoshiCodeFormer: ExplainedIf you have used stable diffusion then I guess you must have come across this paper. In this article I have done my best to explain it.Sep 14, 2023Sep 14, 2023
Utkarsh DoshiAutomatic Captcha SolverThis article describes the machine learning aspect of a Chrome extension that solves captcha on my college’s course registration website.Aug 30, 2023Aug 30, 2023
Utkarsh DoshiT2I adapter: What is it? and How does it work?t2i models have taken the world by storm and these adapters are just supporting itJul 28, 2023Jul 28, 2023