Computer Vision for Geosciences

from classical methods to modern approaches using Deep Learning

Home

2026 course edition

Third course edition, taught at UNAM in 2026.

Note: This edition puts more emphasis on Deep Learning, including group projects, at the expense of some Computer Vision and Machine Learning content covered in prior editions. The following topics, covered in 2024, were not taught this year: CV3 Morphology & Segmentation (slides), CV4 Homography (slides), CV5 Features & Motion Estimation (slides part 1, slides part 2), ML3 Principal Component Analysis (slides), ML4 Random Forests (slides).

Instructors

Lectures

Introduction

Lecture 1: Python & Jupyter crash course

Lecture 2 (CV1): Digital image basics

Lecture 3 (CV2): Filtering

Lecture 4 (ML1): Regression

Lecture 5 (ML2): Classification

Lecture 6 (DL1): Shallow Neural Networks (MLPs)

Lecture 7 (DL2): Convolutional Neural Networks (CNNs)

Lecture 8 (DL3): Convolutional Encoder-Decoder achitectures

Projects

Group project (Weeks 11-14), in collaboration with Sophie Giffard-Roisin (IRD - ISTerre)