Posts by Collection

Free Space Optical Channel Estimation Based on Deep Learning Algorithms

Published in 2023 International Workshop on Fiber Optics on Access Networks (FOAN), 2023

This work introduces a channel estimation scheme based on deep learning algorithms (DNN, CNN, RNN) for Free Space Optical communication systems, outperforming traditional estimation techniques.

Recommended citation: P. Ndiaye, A. Sow, M. Diop, I. Diop. (2023). "Free Space Optical Channel Estimation Based on Deep Learning Algorithms." 2023 International Workshop on Fiber Optics on Access Networks (FOAN), IEEE, pp. 27-31. https://scholar.google.com/citations?view_op=view_citation&hl=en&user=3dqXpXMAAAAJ&citation_for_view=3dqXpXMAAAAJ:TQgYirikUcIC

Multi UAVs Preflight Planning in a Shared and Dynamic Airspace

Published in Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS), 2026

DTAPP-IICR: a Delivery-Time Aware Prioritized Planning method with Incremental and Iterative Conflict Resolution for large-scale UAV fleets in dynamic, shared airspace.

Recommended citation: A. Sow, M. Rodriguez Cesen, F. M. C. de Oliveira, M. Wzorek, D. de Leng, M. Tiger, F. Heintz, C. E. Rothenberg. (2026). "Multi UAVs Preflight Planning in a Shared and Dynamic Airspace." Proceedings of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026). https://arxiv.org/abs/2602.12055

Computer Vision course

Master1 class, Dakar Institute of Technology(DIT), 2023

Designed and delivered a Master’s level course on computer vision covering fundamental image processing techniques and state-of-the-art deep learning methods. The curriculum included image filtering and edge detection, introduction to Convolutional Neural Networks (CNNs), PyTorch tutorials, image classification, object detection, image segmentation (semantic and instance segmentation), optical flow, and action recognition. Students learned to implement and train deep learning models for various computer vision tasks, gaining practical experience with modern frameworks and datasets. The course combined theoretical foundations with hands-on projects, enabling students to develop real-world computer vision applications.

Reinforcement Learning course

Master2 class, Dakar Institute of Technology(DIT), 2023

Taught a comprehensive Master’s level course covering fundamental concepts and advanced techniques in reinforcement learning. The course included theoretical foundations of Markov Decision Processes (MDPs), value-based methods (Q-learning, DQN), policy-based methods (REINFORCE, Actor-Critic), and deep reinforcement learning algorithms. Students gained hands-on experience implementing RL algorithms using TensorFlow and PyTorch, working on classic environments like CartPole and MountainCar, as well as more complex tasks. The course emphasized both theoretical understanding and practical implementation skills, preparing students for research and industry applications in autonomous systems, robotics, and intelligent decision-making.

TDDE13 - Multiagent Systems Permalink

Teaching Assistant, Linköping University, 2024

Responsible for laboratory sessions and seminars in this advanced course on multi-agent systems. The course covers agentic AI with hands-on labs on building LLM agents using local open-source models (Ollama with llama3.2) and multi-agent communication and coordination. Additionally, I supervise multi-agent reinforcement learning laboratory sessions where students implement and experiment with MARL algorithms. I also facilitate seminars on agents and game theory, mechanism design, auctions, and social choice, guiding students through exercise sets and providing feedback on their individual research reports. The course emphasizes both theoretical foundations of multi-agent systems and practical implementation skills, preparing students for research and industry applications in collaborative AI systems.

TDDE78 - Reinforcement Learning Permalink

Lab Responsible, Linköping University, 2024

Responsible for laboratory sessions in this 6-credit advanced course on reinforcement learning. I design and supervise hands-on laboratory assignments based on scientific literature in reinforcement learning, covering fundamental concepts and state-of-the-art algorithms. Students work on practical implementations of RL algorithms, gaining experience with both theoretical understanding and practical skills. The labs focus on research-oriented projects that help students develop capabilities in planning, conducting, and evaluating research and development projects in the field of reinforcement learning.