Cédric Manouan

Hello from the other side! I am a Deep Learning Engineer who loves building Neural Networks of all kinds.🤗
I previously served as a Research Assisociate within the AI & Robotics Lab at Carnegie Mellon University Africa, which is affiliated with the College of Engineering at Carnegie Mellon University. Under the supervision of Prof. David Vernon, I was engaged in research focused on the practical applications of Natural Language Processing within the field of cognitive robotics.

I hold an Master's degree in Information Technology (Applied Machine Learning) from Carnegie Mellon University, where I was fortunate to work as a research assistant for Dr. Moise Busogi (also a teaching assistant for his 04-800 AB ML4EO course) and a teaching assistant for Prof. Bhiksha Raj (11-785 Introduction to Deep Learning) in Fall 2022 & Spring 2023.
Prior to joining CMU, I completed my BSc. in Systems, Computer Networks and Telecommunications (SRIT) at ESATIC, where I worked in the Innovation and Development Unit (CID) reporting to Dr. Ghislain Pandry and Dr. Franklin Kouassi.

Among other things, I am particularly interested in AI Policy, especially in the African context.

When I am not writing computer programs, you'd find me somewhere hyping Cote d'Ivoire for being the 2023 AFCON Champions ⭐⭐🌟, or arguing about NBA players (more about Victor Wembanyama these days), and Ethics in AI.
I do enjoy playing basketball and take random (night) walks.

Email / GitHub / CV Eng. - CV Fr. / Twitter / LinkedIn

Research Interests & Deep Learning Engineering

I have a broad interest in leveraging NLP techniques to address communication barriers and access to high-quality knowledge in the context of African languages.
I am also interested in developing autonomous agents that can engage with humans using natural language.
My research activities, to date, have revolved around the implementation of ML/DL models for various applications, such as speech recognition, text and image classification, and machine learning for Earth Observation (ML4EO). You can find some of my work in the following sections.

My most recent assignments involved exploring the utilization of robot control networks and approaches from Vision-Language-Action (VLA) modeling for semantically modulating Joint Episodic-Procedural Associative Working Memories. My primary objective was to create a Behavioral Episode model referred to as S-JEP, implementing the working memory which is at the core of the Situation Model Framework.

Recently, I have been delving into exploring the concepts of "System 2 Deep Learning," and "Autonomous Machine Intelligence" inspired by ideas presented by Prof. Yann Lecun (see the corresponding paper and his talk here) and Prof. Yoshua Bengio (you can watch one of his talks on the topic here). It's worth noting that both of these ideas expand upon the concept of the Dual-system theory, which was popularized by Daniel Kahneman in Thinking Fast and Slow (Talk/Book).
My hypothesis in this regard is being complied as part of an ongoing independent research titled Language is Almost All you Need (preprint coming soon).

I have a particular affinity for subjects related to Leading and Managing Technological Innovation. My enthusiasm for this area intensified following my enrollment in a course of the same name, instructed by Prof. Mark Kryder. In the context of African AI policy, I am keenly interested in examining the performance of our countries across five (5) key axes:

Algorithms (ML/DL)

Data (Text, Images, Speech, etc.)

Compute infrastructures (Effective supercomputing power...compute (GPU) clusters essentially)

Workforce and Key Players/Contributors (AI specialists; Disseminetion levels)

Regulations (Safety, Ethics, Security & data privacy).