Cheick Tidiani Cissé
Research Scientist, Generative AI
Belfort, Francetidianicisse@gmail.comcheicktidianicisse.comlinkedin.com/in/cheick-tidiani-cissé-ph-d-83364613agithub.com/Tysby101
Profile
I build generative models that leave the lab — diffusion and GANs for radio maps, transformers for low-resource translation, sequence models for human mobility.
PhD in Computer Science (CIFRE, Orange Labs × UTBM). 8 peer-reviewed papers including a sole-authored paper at Findings of ACL 2026, 1 patent filed, Top-5 at the ICASSP 2025 Grand Challenge. Reviewer for NeurIPS and AAAI. CNU-qualified, sections 27 and 61.
Experience
Applied AI Scientist · Orange Business
Apr 2025 — Present- Research lead on transformer-based generative models that synthesise realistic mobility traces, letting Orange derive value from operator data without exposing real subscriber trajectories.
- Engineered the data pipelines behind it: Apache Airflow orchestrating distributed processing across multiple hosts, parallelised horizontally with Docker and Celery.
- Took models from notebook to production: reproducible training, tracked experiments, containerised serving, MLOps practice throughout.
- Defined and opened a Master’s-level research internship on generative mobility modelling; scoped the subject and supervised the work.
PhD Researcher — Deep Learning for Radio Planning · Orange
Dec 2021 — Dec 2024- Designed IRGAN, a conditional GAN that predicts a full indoor radio map from a single 2D floor plan at 2.5 dB RMSE — fast enough to sit inside an optimisation loop, unlike ray tracing.
- Extended the approach with conditional diffusion models and a physics-informed ResUNet, placing Top-5 of 16+ teams in the ICASSP 2025 Grand Challenge on generalisable path-loss prediction.
- Solved the sim-to-real gap by fine-tuning on crowd-sourced 5 GHz measurements from live home networks (EuCAP 2024).
- Coupled the learned propagation model with metaheuristics — simulated annealing, tabu search, genetic algorithms — for automatic access-point placement under coverage and throughput constraints.
Research Intern — Remote Sensing · LISIC, Université du Littoral Côte d’Opale
Mar 2021 — Sept 2021- Designed a fusion architecture from scratch that uses Sentinel-3 hyperspectral data to fill gaps in Sentinel-2 multispectral time series, raising effective temporal resolution.
- Published the result at ICASSP 2022 and GRETSI 2022 as first author.
Selected publications
- 2026 Kumatigi: Quality-Driven Data Augmentation for Low-Resource Machine Translation· Sole author — Findings of ACL 2026
- 2025 Generalizable Indoor Path Loss Prediction — ICASSP 2025
- 2024 Fine Tuning an AI-based Indoor Radio Propagation Model with Crowd-sourced Data — EuCAP 2024
- 2023 IRGAN: cGAN-based Indoor Radio Map Prediction — IFIP Networking 2023
- 2022 A New Deep Learning Method for Multispectral Image Time Series Completion Using Hyperspectral Data — ICASSP 2022
+ 3 national symposium papers (ROADEF, GRETSI)
Peer review & qualifications
- Reviewer — AAAI 2027, NeurIPS 2026
- CNU qualification, sections 27 & 61 — Nationally qualified for Associate Professor positions in Computer Science and Signal Processing.
- Patent (filed) — Incorporating propagation-model error into indoor Wi-Fi optimisation.
Education
PhD in Computer Science — Université de Technologie de Belfort-Montbéliard (UTBM) — FEMTO-ST
2021–2024MSc, Telecommunications & Network Engineering — France
2019–2021Classes préparatoires — Mathematics, Physics, CS — Tétouan, Morocco
2016–2018Skills
- Generative modelling:
- Diffusion models, GANs / cGANs, VAEs, Transformers, Sequence models
- Machine learning:
- PyTorch, TensorFlow, Hugging Face, scikit-learn, Transfer learning, Architecture search
- Applied domains:
- NLP / low-resource MT, Computer vision, Remote sensing, Radio propagation, Combinatorial optimisation
- Engineering & MLOps:
- Python, Docker, MLflow, Airflow, GCP, Slurm, Linux / Bash, Git, C, TypeScript
Languages
- FrenchNative
- BambaraNative
- EnglishProfessional