Greetings!
I’m a Research Engineer in Huawei Noah’s Ark Lab at Montreal, working in the NetMind team. Previously, I completed a master’s degree in Computer Science at McGill University and did my undergrad in Computer Engineering at Sharif University of Technology. In my high school years I had the opportunity to participate in Internationl Physics Olympiad (IPhO) competitions which cemented in me a love for physics and the realm of theoretical thinking
Experience
AI Expert @ FANAP
I’ve had the chance to work alongside many talented people and learn a lot from them. I started my career as an intern at FANAP (Tehran, Iran) as an AI expert. I directly contributed to 2 projects. During that time I had the chance to explore my interest in the field of Artificial Intelligence (AI) and Machine Learning (ML). The first project involved creating an Automated Speech Recognition (ASR) system in persian language. This model was used as an assistant to help people with visual disabilities to complete their mobile banking tasks. The goal of the second project was to create a smart gym mirror for home workouts that would give helpful feedback on how to perform workouts. I helped in creating and compressing the pose estimation model and to deploy it on mobile phones with the help of Unity engine.
Associate Researcher @ Huawei Noah's Ark Lab
Currently, I am working in Huawei Noah’s Ark Lab as a research engineer. At Net-Mind we focus on developing the necessary tools for next stage Native-AI in 6G. With the help of my colleagues we have implemented, tested and proposed different methods for decentralized learning and knowledge aggregation of ML models in distributed environments. As a branch of this project I am currently researching on probabilistic and information theoretic methods to measure, constrain and direct the flow of ML Information in networks to optimize the communication of knowledge and information. My main focus is on Optimal Transport, Information Bottleneck and various other methods branching from these research directions.
Research Experience
Information Network Learning (INL) Lab @ Sharif University of Technology
In my 2nd and 3rd year of undergraduate I started working on two projects at INL lab to get more in touch with the relevant research directions in AI and ML. For my first project we developed a model that uses Polynomial Codes to solve a distributed SVM problem. Our method achieved a 8x communication reduction on Amazon EC2 nodes. For the 2nd project we developed a Variational Auto-Encoder (VAE) model that sends compressed images on a noisy network and decompresses them on the receiver side. We intuitively and empirically showed that this method achieved better results in noisier environments compared to previous state-of-the-art methods. This paper was published at IST 2020.
Advanced Network Research Lab (ANRL) @ McGill University
The focus of my Master’s thesis was on predicting and adapting to data distribution drifts in large distributed networks. We developed a triple attention model that keeps local models in a network as robust as possible to data drifts meanwhile minimizing the communication necessary to adapt to these issues. This paper is on Arxiv.
Personal Projects
During the years through my own curiosity, I have started multiple projects and played around with different tools and methods. You can find the codes and documentations for my projects in my github page. My most recent project, was a peer-to-peer (P2P) cloud system that connects gamers to each other (G2G) that can be deployed in a blockchain and supports automatic payment. This project is still ongoing. I also created a telegram bot that lets user register and connect to a decentralized lottery on the Ethereum blockchain. This project was written in Solidity and supports payments in USDT.
Contact
I am curious about anything related to physics, probability theory, and machine learning! Let's get in touch.
Feel free to send me an email at ahns3.ahe@gmail.com.