Downloads
Training Materials & Resources
Overview
Download training materials for the CoPhil EO AI/ML Training Programme. All materials are provided under open licenses for educational use.
Day 1: EO Data, AI/ML Fundamentals & Geospatial Python
📹 Day 1 Full Session Recording
Duration: 6h 53m | Size: 1.19 GB | Format: MP4 1080p
Session 1: Copernicus Sentinel Data & Philippine EO Ecosystem
Presentation: View Online Download PDF
Topics: - Copernicus Programme overview - Sentinel-1 SAR and Sentinel-2 optical missions - Philippine EO agencies (PhilSA, NAMRIA, DOST-ASTI, PAGASA) - CoPhil Mirror Site and infrastructure
Session 2: AI/ML Fundamentals for Earth Observation
Presentation: View Online Download PDF
Topics: - What is AI/ML and the EO workflow - Supervised vs. Unsupervised learning - Introduction to neural networks and CNNs - Data-centric AI paradigm
Session 3: Python for Geospatial Data
Presentation: View Online Download PDF
Jupyter Notebook: Download Notebook Open in Colab
Topics: - Google Colab setup - GeoPandas for vector data - Rasterio for raster data - Coordinate reference systems - Philippine case study: Palawan land cover
Session 4: Introduction to Google Earth Engine
Presentation: View Online Download PDF
Jupyter Notebook: Download Notebook Open in Colab
Topics: - Earth Engine authentication and initialization - ImageCollection filtering - Sentinel-1 SAR and Sentinel-2 optical data access - Cloud masking and temporal compositing - Philippine case study: Metro Manila monitoring
Day 2: Machine Learning for Land Cover Classification
📹 Day 2 Full Session Recording
Duration: 6h 05m | Size: 875.1 MB | Format: MP4 1080p
Session 1: Random Forest Classification
Presentation: View Online Download PDF
Jupyter Notebooks: Theory Notebook Hands-on Lab Open Theory in Colab Open Lab in Colab
Topics: - Decision trees and ensemble methods - Random Forest algorithm - Feature importance and model interpretation - Land cover classification with Sentinel-2
Session 2: Palawan Land Cover Lab
Presentation: View Online Download PDF
Jupyter Notebook: Extended Lab Open in Colab
Topics: - Model evaluation and validation - Hyperparameter tuning - Cross-validation strategies - Handling imbalanced datasets
Session 3: Deep Learning Fundamentals
Presentation: View Online Download PDF
Jupyter Notebook: Theory Interactive Open in Colab
Topics: - Neural networks architecture - Backpropagation and optimization - Introduction to PyTorch/TensorFlow - Building simple neural networks
Session 4: Convolutional Neural Networks for EO
Presentation: View Online Download PDF
Jupyter Notebooks: CNN Classification Transfer Learning Open CNN in Colab Open Transfer Learning in Colab
Topics: - CNN architecture and components - Transfer learning for EO - Image classification with CNNs - Philippine land use case studies
Day 3: Semantic Segmentation & Object Detection
📹 Day 3 Full Session Recording
Duration: 4h 52m | Size: 610.8 MB | Format: MP4 1080p
Session 1: U-Net Semantic Segmentation
Presentation: View Online Download PDF
Topics: - Semantic segmentation fundamentals - U-Net architecture (encoder, decoder, skip connections) - Loss functions for segmentation (Cross-Entropy, Dice, IoU, Combined) - EO applications (flood mapping, land cover, buildings, roads) - Philippine case studies (Typhoon Ulysses, Palawan)
Session 2: Flood Mapping with U-Net
Presentation: View Online Download PDF
Jupyter Notebook: Flood Mapping Lab Open in Colab
Topics: - SAR data for flood detection - U-Net implementation - Training and validation - Philippine flood mapping case study
Session 3: Object Detection Theory
Presentation: View Online Download PDF
Topics: - Object detection frameworks - YOLO and Faster R-CNN - Detection vs segmentation - Evaluation metrics
Session 4: Object Detection Lab
Presentation: View Online Download PDF
Jupyter Notebook: Object Detection Lab Open in Colab
Topics: - Transfer learning for object detection - Pre-trained models (SSD MobileNet, Faster R-CNN, YOLO) - Building detection from Sentinel-2 imagery - mAP evaluation metrics - Metro Manila urban monitoring case study - Operational deployment pipeline
Day 4: Time Series Analysis & Advanced Topics
📹 Day 4 Full Session Recording
Duration: 5h 01m | Size: 486.3 MB | Format: MP4 1080p
Session 1: LSTM for Time Series
Presentation: View Online Download PDF
Jupyter Notebook: LSTM Demo Open in Colab
Topics: - Why time series analysis matters for EO - Recurrent neural networks and vanishing gradients - LSTM architecture (gates, cell state, memory) - Philippine seasonal patterns and drought forecasting - Mindanao case study with Sentinel-2 NDVI - Multivariate time series modeling
Session 2: Drought Monitoring with LSTM
Presentation: View Online Download PDF
Jupyter Notebook: Drought Lab Open in Colab
Instructor Notebook: Download Notebook
Topics: - Multi-variate time series - LSTM implementation for drought - Feature engineering - Philippine drought case study
Session 3: Emerging AI Technologies
Presentation: View Online Download PDF
Topics: - Self-supervised learning - Vision transformers - Foundation models overview - Prithvi and other EO models
Session 4: Synthesis & Best Practices
Presentation: View Online Download PDF
Topics: - End-to-end EO AI/ML workflow - Model deployment - Operational considerations - Future directions in EO AI/ML
Cheat Sheets (PDF)
Quick reference guides for common operations:
Python Basics
Essential Python syntax and operations
GeoPandas Reference
Vector data operations
Rasterio Commands
Raster data handling
Earth Engine API
GEE Python commands
Sentinel Missions
Band specifications
Spectral Indices
Common formulas (NDVI, NDWI, etc.)
Additional Resources
Recommended Reading
Earth Observation: - Copernicus Open Access Hub User Guide - Sentinel-1 Toolbox Documentation - Sentinel-2 User Handbook
Python for Geospatial: - GeoPandas User Guide - Rasterio Quickstart - Python Geospatial Development
Google Earth Engine: - Earth Engine Python API Guide - Earth Engine Community Tutorials - Awesome Earth Engine
AI/ML for EO: - Deep Learning for Earth Observation (Springer) - Fundamentals of Machine Learning for Predictive Data Analytics - Data-Centric AI Resource Hub
Support & Feedback
Having Issues?
- Check the FAQ for common problems
- Review the Setup Guide for installation help
- Contact training coordinators: skotsopoulos@neuralio.ai
Suggest Improvements
We welcome your feedback to improve the training materials:
- Report broken links or errors
- Request additional materials
- Share feedback on content
Contact: skotsopoulos@neuralio.ai
All materials are regularly updated. Check back for new resources and improved versions.