logo_X
Logo_LIX
Logo_SX
Profile

Simo Alami

Ph.D. Candidate · @Ecole Polytechnique · Lab: LIX · mohamed.alami-chehboune@polytechnique.edu

I am a Ph.D. candidate in LIX at Ecole Polytechnique where I am advised by Prof. Jesse Read. Prior to that, I studied Mathematics and Computer Science at Université Pierre et Marie Curie and Ecole CentraleSupélec.

My research interests include Deep learning, Metric Learning and Reinforcement Leaning. More specifically, I am interested in studying algorithms that learn from experience using self learned metrics.

My main area of study is Inverse Reinforcement Learning as a tool to teach robots logical concepts they could leverage to perform several tasks even when they never encountered them before.

Keywords: Deep Learning, Inverse Reinforcement Learning, Metric Learning, Meta-Learning


News

[02/22] Our paper on Zero-Shot Clustering through Metric Transfer Learning  has got accepted to IDA 2023.

[09/22] We presented our work realized during my stay in Accenta on Non-Invasive Load Monitoring at ECML 2022 in Grenoble, France.

[08/22] I presented our work on Curiosity augmented Metropolis-Hastings method for Optimal Policies Distributions at EUSIPCO 2022 in Belgrade, Serbia. Slides are available here.

[04/22] I have got awarded with a grant from Nvidia to support my research on Meta-Inverse Reinforcement Learning.

[02/22] I began my gap semester as a visiting researcher at Accenta.

[03/21] I had the honour to organize the Electromobilité et territoires webinar at IRT SystemX

[02/21] I won the 3rd Prize of Hi Paris Hackathon on Smart Grid Energy management using a RL Solution

[12/19] I presented and published My Master thesis on Deep Learning Methods for 6 DoF Pose Estimation. The correponding report can be found here.

[04/19] I've been honoured to join Inria's Magrit team for a 6 months internship


Teaching

Advanced Machine Learning and Autonomous Agents: Computer Science, Ingénieur 3A/Master1, Ecole Polytechnique. Winter 21, 22, 23. In charge of tutorials, lectures taught by Pr.Jesse Read

2020 - Present

Publications

Transferable Deep Metric Learning for Clustering
Simo Alami, Rim Kaddah, and Jesse Read.
Symposium of Intelligent Data Analysis (IDA), 2023
[pdf]

Conv-NILM-Net, a Causal and Multi-appliance Model for Energy Source Separation
Simo Alami, Jérémie Decock, Rim Kaddah, and Jesse Read.
Joint European Conference on Machine Learning and Knowledge Discovery in Databases (ECML), 2022
[pdf]

CAMEO: Curiosity Augmented Metropolis for Exploratory Optimal Policies
Simo Alami, Fernando Llorente, Rim Kaddah, Luca Martino and Jesse Read.
European Signal Processing Conference (EUSIPCO), 2022
[pdf]

Deep learning Methods For 6DoF Pose Estimation
Master Thesis.
Simo Alami
[pdf]


Interests

I am fond of robotics and highly interested in the emerging field of Soft Robotics. During my free time, I enjoy cinema and playing guitar. Some of my guitars are displayed below and my tunes can be found on my soundcloud.

Lag Arkane 1000, Old Port

Epiphone Les Paul Modern, Sparkling Burgundy

Satch Blues, Personnal Composition

Canon Rock, Johann Pachelbel, arranged for guitar by Jerry C

The Essence of the Blues, Personnal Composition

Starry Night, Joe Satriani

Ten Words, Joe Satriani