Othmane Echchabi

About Me

I’m Othmane Echchabi, an incoming M.Sc. student in Computer Science at McGill University / Mila Québec AI Institute. My interests centre on computer vision, self-supervised learning and trustworthy ML, especially for climate- and sustainability-focused applications. I’ve led a range of applied machine-learning projects across remote sensing, mobility analytics and environmental monitoring—highlights are in my CV and the Projects section of this site. When I’m not coding, you’ll catch me playing football, or watching football, or on a trip, or planning a trip.

Education

Current Projects

Here are a selection of my recent projects spanning satellite‑based environmental monitoring, machine learning for mobility and carbon footprint prediction, and AI‑driven computer vision applications.

←

Tracking Progress Towards SDG6

Assessed piped water and sewage access in 50 African countries using vision transformers on satellite imagery (97% accuracy), and built a comprehensive dataset to monitor Sustainable Development Goal 6.

Fertilizer Recommendation System

Built a global dataset integrating a decade of crop yield and fertilizer data and created an ML pipeline for field fertilizer recommendations. Co-created a decision tool with an integrated LLM (RAG).

Carbon Clever & Trip Mode Prediction

Developed a transformer‑based self‑supervised model achieving 92% accuracy on GeoLife and 95% on MOBIS for trip‑mode identification, and designed CarbonClever to help users reduce their individual carbon footprints.

Monitoring Spartina Alterniflora

Applied self‑supervised learning with ViTs to identify and map the distribution of saltmarsh cordgrass from coastal wetland imagery, presented at Duke’s Climate+ Symposium.

Assessing Climate Change Risk

Created a geospatial database in ArcGIS and used NLP on 1,000+ articles to map saltwater intrusion and sea level rise risks for rural coastal plain communities.

Yancheng Wetlands Degradation

Conducted a 25‑year spatiotemporal NDVI analysis of Yancheng coastal wetlands degradation and developed a state‑of‑the‑art wetland segmentation model.

DigCai Database

Collaborated as an editor on the first release of DigCai v4.0.0: a TEI-XML marked-up corpus of Cairo’s urban development articles (1828–1914) from the Egyptian Government Gazette, in Arabic and Ottoman Turkish, supported by an NEH grant.

→

Publications

• Echchabi, O.; Lahlou, A.; Talty, N.; Manto, J. M.; Lam, K. L.* Tracking Progress Towards Sustainable Development Goal 6 Using Satellite Imagery. Pre-print, 2025. arXiv
• Zhang, Y.; Echchabi, O.; Feng, T.; Zhang, W.; Liao, H.-K.; Chang, C.* Predicting Human Mobility Using Dense Smartphone GPS Trajectories and Transformer Models. Under review, 2025. PDF
• Mestyan, A.; Cayless, H.; Volait, M.; et al. (eds.); Echchabi, O. (ed.) DigCai Database v4.0.0: A TEI-XML Corpus of Cairo’s Urban-Development Articles (1828–1914). Zenodo 15255721, 2024. Zenodo

For a complete list, see my CV.

Teaching

Duke Kunshan University

COMPSCI101: Introduction to Computer Science (Teaching Assistant)
COMSPCI201: Introduction to Programming and Data Structures (Teaching Assistant)
MATH105: Calculus (Teaching Assistant)
MATH201: Multivariable Calculus (Teaching Assistant)
MATH202: Linear Algebra (Teaching Assistant)
MATH317: Quantitative Finance (Teaching Assistant)

Duke University

MATH201: Multivariable Calculus (Teaching Assistant)

Curriculum Vitae

For a detailed overview of my academic and professional background, please find my CV below.

• Download CV
• View Online

• © Untitled. All rights reserved.
• Design: HTML5 UP