Jash Patel Email & Phone Number

Bengaluru, Karnataka, India

As a Machine Learning Researcher at Ayna AI, I specialized in diffusion models with a focus on virtual try-on applications. My primary contributions included:- Training and developing custom diffusion models tailored for virtual try-on tasks, exploring various architectures to optimize performance.- Creating and curating a proprietary dataset, which involved generating captions, creating masked images and validating the dataset to establish in-house quality criteria.- Investigating vision-language models, particularly the CLIP model, to enhance the validation process for the captions generated from our dataset, ensuring alignment with visual content.

Bengaluru, Karnataka, India

During my internship at Ayna AI, I focused on transformer-based diffusion models, specifically the Diffusion Transformer (DiT) model. My key responsibilities included:- Conducting in-depth studies on various diffusion models, including Pix-Art Alpha, Sigma, Delta, and ControlNet, to understand their architectures and applications.- Exploring the behavior of Stable Diffusion 3 and comparing its performance with other DiT-based architectures.

Jodhpur, Rajasthan, India

- Contributed to cutting-edge research at the Image Analysis and Biometrics Lab in the Computer Science and Engineering Department, specializing in image-based generative models such as GANs and DDPMs.- Engaged in projects involving Vision Transformers (VIT) for small datasets

Surat, Gujarat, India

- Designed strong and scalable backend systems with Python and Django, ensuring optimal performance. - Led frontend development using React.js and TypeScript to create user-friendly interfaces for an improved user experience.

Master Of Technology - Mtech, Artificial Intelligence, 7.77

Bachelor Of Engineering - Be, Computer Engineering, 9.45