Presented with short summaries — In my previous post on Medium, I have discussed four Deep Learning papers from late 2021 that I found very disruptive. These four papers were MetaFormer (Paper and Code), Understanding AlphaZero (Paper), 3D Style GAN (Paper and Code), and SeMask (Paper and Code). Since I noticed that the post had…

A straightforward introduction to Vector Quantized Generative Adversarial Networks Some days ago, I was trying to understand a brand new architecture, MaskGit, a Masked Generative Image Transformer based on VQ-GAN. …

Presented with short summaries — Recently, I have started collaborating with the team of MarkTechPost, a California-based Artificial Intelligence news platform. My role is to choose two papers each month, which I consider fundamental, and write a short, easy-to-understand summary. I have decided to post these summaries (to be honest, a summary of these summaries)…

A brief resume of two years of my primary Ph.D. research — Table of contents Introduction Background The full pipeline of our system Results Conclusion References 1. Introduction Since the beginning of my Ph.D., I have been collaborating with the orthopedic team of the CTO (Center for Orthopaedic Trauma) of Turin (Italy), to develop an algorithm able to assist physicians in fracture diagnosis. We select the femur…

A user-friendly explanation of the brand new architecture introduced by Google — Table of contents Introduction The basis of CNNs From Self-Attention to Vision Transformers CoAtNet Conclusion References 1. Introduction On September 15, 2021, a new architecture defined the state-of-the-art (SOTA) in the ImageNet competition. CoAtNet (pronounced “coat” net) achieved a 90.88% top-1 accuracy with the huge JFT-3B dataset. …