Yoni Gottesman

Medical records often contain scanned text, whether they are physical documents scanned into the system or digital documents from which text cannot be extracted.

Looking at your agent run for the first time is a truly magical experience.

TotalSegmentator is a tool for medical segmentation of most major anatomical structures.

These days, fine-tuning large language models (LLMs) seems to be getting easier and easier.

Generative models, such as ChatGPT and Llama2, excel in question-answering tasks given a specific context.

Some time ago, with the explosion of DALLE2 and stable diffusion, I also wanted to understand how these models tick, so I opened the Denoising Diffusion Probabilistic Models (DDPM) paper.

In this post, I will use a vision transformer to classify ECG signals and use the attention scores to interpret what part of the signal the model is focusing on.

In recent years, self-supervised learning for computer vision has transitioned from pretext tasks to siamese based representation learning.

If you are working in a big tech company, chances are you have a CI/CD pipeline built by engineers and devops for you.

Conditional GAN architectures such as cGAN learn to generate fake images with a specific condition.