add set up via upload

Author: elenaivadreyer

Makefile

@@ -0,0 +1,36 @@
+SHELL := /bin/bash
+
+.PHONY: create_venv install install-torch lint help
+
+help:
+	@echo "Available targets:"
+	@echo "  create_venv    - Create virtual environment and install dependencies (without torch)"
+	@echo "  install        - Install package with dev dependencies (without torch)"
+	@echo "  install-torch  - Install PyTorch CPU-only version"
+	@echo "  lint           - Run pre-commit hooks on all files"
+
+create_venv:
+	@echo "Creating virtual environment..."
+	python3 -m venv .venv
+	@echo "Installing dependencies (without torch)..."
+	.venv/bin/python3 -m pip install --upgrade pip
+	.venv/bin/pip install -e .[dev]
+	@echo ""
+	@echo "Virtual environment created successfully!"
+	@echo "To install PyTorch (CPU-only), run: make install-torch"
+	@echo "Or manually install with:"
+	@echo "  pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu"
+
+install:
+	python3 -m pip install --upgrade pip
+	pip install -e .[dev]
+	@echo "Dependencies installed (without torch)."
+	@echo "To install PyTorch (CPU-only), run: make install-torch"
+
+install-torch:
+	@echo "Installing PyTorch CPU-only version..."
+	pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
+	@echo "PyTorch installed successfully!"
+
+lint:
+	pre-commit run --all-files

README.md

@@ -1,37 +1,98 @@
-[![Review Assignment Due Date](https://classroom.github.com/assets/deadline-readme-button-22041afd0340ce965d47ae6ef1cefeee28c7c493a6346c4f15d667ab976d596c.svg)](https://classroom.github.com/a/15CTOpCR)
-# Deep Learning E1394 - Tutorial
-
-## Submitting your work
-
-Once the deadline has passed, you are not able to push to your repository anymore or update it in any other way and all changes on the main branch will be considered for grading. In other words, there is no "submit" button which you need to press. You just need to ensure that all your work has been pushed / uploaded to GitHub before the deadline.
-
-Uploading a new or edited file to GitHub involves three steps:
-1. Add the file to the staging area:
-    ```
-    git add <file to be added>
-    ```
-2. Create a commit:
-    ```
-    git commit -m <commit msg - briefly describe your change>
-    ```
-3. Push the local commit to GitHub:
-    ```
-    git push
-    ```
-
-To avoid conflicts when multiple people are working on the same file, we recommend to push your changes as frequently as possible and always pull the latest changes from GitHub before you start working. Also, as a nice teammate, test your changes before you push them and use meaninful commit messages to make it easy to understand your changes.
-
-## Asking for help
-
-If you have important questions or believe to have spotted an error, please open an issue on your repository and mention the teaching assistant with @chiara-fb and @henrycgbaker.
-
-## Useful Links
-* Git best practices: https://gist.github.com/luismts/495d982e8c5b1a0ced4a57cf3d93cf60
-* How to write a good git commit message: https://cbea.ms/git-commit/
-* Book "Dive into Deep Learning": https://d2l.ai/
-* Tutorial on pushing and submitting work with GitHub classroom: https://www.youtube.com/watch?v=jXpT8eOzzCM
-* Neural network from scratch:
-    * https://towardsdatascience.com/math-neural-network-from-scratch-in-python-d6da9f29ce65
-    * https://pythonalgos.com/create-a-neural-network-from-scratch-in-python-3/
-    * https://github.com/casperbh96/Neural-Network-From-Scratch/blob/master/NN_From_Scratch.ipynb
-    * https://www.codingame.com/playgrounds/59631/neural-network-xor-example-from-scratch-no-libs
+# Few-Shot Learning for Rooftop Detection in Satellite Imagery
+
+## Tutorial Overview
+
+This tutorial demonstrates few-shot learning techniques for semantic segmentation of satellite imagery. The dataset contains high-resolution satellite images of Geneva, Switzerland, with corresponding segmentation labels for rooftop detection.
+
+📓 **[View Tutorial Notebook (HTML)](docs/tutorial_few_shot_learning.html)**
+
+### Learning Outcomes
+- Understanding few-shot learning concepts for image segmentation
+- Working with satellite imagery and segmentation masks
+- Implementing and evaluating few-shot learning models for rooftop detection
+
+## Video Tutorial
+
+<!-- VIDEO PLACEHOLDER: Replace the link below with your tutorial video -->
+[![Tutorial Video](https://img.youtube.com/vi/VIDEO_ID/0.jpg)](https://www.youtube.com/watch?v=VIDEO_ID)
+
+*Click the image above to watch the tutorial video*
+
+## Quick Start
+
+### Installation
+
+```bash
+# Clone the repository
+git clone https://github.com/elenaivadreyer/tutorial-group-1.git
+cd tutorial-group-1
+
+# Create virtual environment
+python3 -m venv .venv
+source .venv/bin/activate
+
+# Install the package (without PyTorch)
+pip install -e .
+
+# Install PyTorch (CPU-only version recommended to save disk space)
+# CPU-only: ~730MB vs CUDA version: ~7GB
+pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
+
+# Or use the Makefile
+make install
+make install-torch
+```
+
+**Note:** PyTorch is not installed by default due to its large size. The CPU-only version is recommended for most use cases and saves significant disk space.
+
+### Running the Tutorial
+
+Open the main tutorial notebook:
+```bash
+jupyter notebook notebooks/tutorial_few_shot_learning.ipynb
+```
+
+## Project Structure
+
+```
+├── notebooks/
+│   └── tutorial_few_shot_learning.ipynb     # Main tutorial notebook
+├── src/few_shot_utils/
+│   ├── __init__.py                          # Package initialization
+│   ├── data.py                              # Data loading utilities
+│   ├── models.py                            # Model architectures
+│   ├── train.py                             # Training functions
+│   └── evaluate.py                          # Evaluation metrics
+└── docs/                                    # Documentation
+```
+
+## Dataset Description
+
+The dataset consists of:
+- **Satellite Images**: High-resolution RGB satellite images of Geneva, Switzerland
+- **Segmentation Labels**: Binary masks indicating rooftop locations
+- **Resolution**: Images at various resolutions suitable for few-shot learning
+
+## Development
+
+### Linting and Formatting
+
+This project uses [Ruff](https://docs.astral.sh/ruff/) for linting and formatting, including Jupyter notebooks.
+
+```bash
+# Run linting with automatic fixes
+ruff check . --fix
+ruff format .
+
+# Run pre-commit hooks
+pre-commit run --all-files
+```
+
+## References
+
+-
+-
+
+## License
+
+This project is licensed under the MIT License - see the LICENSE file for details.

docs/_config.yml

@@ -0,0 +1,9 @@
+title: Few-Shot Learning Tutorial
+description: Few-shot learning for rooftop detection in satellite imagery
+theme: jekyll-theme-primer
+baseurl: "/tutorial-group-1"
+url: "https://elenaivadreyer.github.io"
+
+# remote_theme: pages-themes/primer@v0.6.0
+# plugins:
+# - jekyll-remote-theme # add this line to the plugins list if you already have one

docs/index.md

@@ -0,0 +1,54 @@
+---
+layout: default
+title: Few-Shot Learning for Rooftop Detection
+---
+
+# Few-Shot Learning for Rooftop Detection in Satellite Imagery
+
+Welcome to the tutorial on few-shot learning for semantic segmentation of satellite imagery, focusing on rooftop detection in Geneva, Switzerland.
+
+## Tutorial Overview
+
+This tutorial demonstrates:
+- Few-shot learning concepts for image segmentation
+- Working with satellite imagery and segmentation masks
+- Implementing and evaluating few-shot learning models for rooftop detection
+
+## Getting Started
+
+### Prerequisites
+- Python 3.11+
+- PyTorch 2.0+
+
+### Installation
+
+```bash
+git clone https://github.com/elenaivadreyer/tutorial-group-1.git
+cd tutorial-group-1
+pip install -e .
+```
+
+### Running the Tutorial
+
+Open the main tutorial notebook:
+```bash
+jupyter notebook notebooks/tutorial_few_shot_learning.ipynb
+```
+
+## Video Tutorial
+
+<!-- VIDEO PLACEHOLDER: Add your video here -->
+*Video coming soon...*
+
+## Dataset
+
+The dataset contains high-resolution satellite imagery of Geneva, Switzerland, with corresponding segmentation labels for rooftop detection.
+
+## Resources
+
+- [GitHub Repository](https://github.com/elenaivadreyer/tutorial-group-1)
+- [Tutorial Notebook](https://github.com/elenaivadreyer/tutorial-group-1/blob/main/notebooks/tutorial_few_shot_learning.ipynb)
+
+## License
+
+This project is licensed under the MIT License.

docs/installation.md

@@ -0,0 +1,111 @@
+# Installation Guide
+
+This guide explains how to set up the tutorial environment.
+
+## Quick Start
+
+### Using Makefile (Recommended)
+
+```bash
+# Create virtual environment and install dependencies
+make create_venv
+
+# Activate virtual environment
+source .venv/bin/activate
+
+# Install PyTorch (CPU-only)
+make install-torch
+```
+
+### Manual Installation
+
+```bash
+# Create and activate virtual environment
+python3 -m venv .venv
+source .venv/bin/activate
+
+# Install package dependencies (without PyTorch)
+pip install -e .
+
+# Install PyTorch CPU-only version
+pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
+```
+
+## Why PyTorch is Optional
+
+PyTorch is not included as a direct dependency because:
+
+1. **Size**: The CUDA version is ~7GB, which can cause disk space issues
+2. **CPU-only alternative**: The CPU-only version is ~730MB and sufficient for this tutorial
+3. **Flexibility**: Users can choose between CPU and GPU versions based on their needs
+
+## PyTorch Installation Options
+
+### CPU-only (Recommended for most users)
+```bash
+pip install torch torchvision --index-url https://download.pytorch.org/whl/cpu
+```
+
+### CUDA (If you have an NVIDIA GPU)
+```bash
+# For CUDA 11.8
+pip install torch torchvision --index-url https://download.pytorch.org/whl/cu118
+
+# For CUDA 12.1
+pip install torch torchvision --index-url https://download.pytorch.org/whl/cu121
+```
+
+For more installation options, see the [official PyTorch installation guide](https://pytorch.org/get-started/locally/).
+
+## Development Setup
+
+### Pre-commit Hooks
+
+This project uses pre-commit hooks for code quality. To set them up:
+
+```bash
+# Install development dependencies
+pip install -e .[dev]
+
+# Install pre-commit hooks
+pre-commit install
+
+# Run hooks manually on all files
+pre-commit run --all-files
+```
+
+### Makefile Targets
+
+- `make help` - Show available targets
+- `make create_venv` - Create virtual environment and install dependencies
+- `make install` - Install package with dev dependencies
+- `make install-torch` - Install PyTorch CPU-only version
+- `make lint` - Run pre-commit hooks on all files
+
+## Devcontainer Setup
+
+If you're using VS Code with devcontainers:
+
+1. Open the repository in VS Code
+2. Click "Reopen in Container" when prompted
+3. The container will build and install dependencies automatically
+4. **Important**: PyTorch is not installed by default. To install it, run:
+   ```bash
+   source .venv/bin/activate && make install-torch
+   ```
+
+## Troubleshooting
+
+### Disk Space Issues
+
+If you encounter disk space issues during installation:
+- Make sure you're using the CPU-only version of PyTorch
+- Clear pip cache: `pip cache purge`
+- Consider using a cloud development environment with more storage
+
+### Import Errors
+
+If you get import errors for torch:
+- Make sure PyTorch is installed: `pip list | grep torch`
+- Verify you're using the correct virtual environment: `which python`
+- Reinstall if necessary: `make install-torch`

docs/memo.md

@@ -0,0 +1,7 @@
+Few-shot learning for satellite image segmentation is crucial for urban planning and disaster response. Traditional deep learning requires thousands of labeled samples, but few-shot learning enables rapid adaptation to new geographic regions with minimal labeled data.
+
+**Policy Applications:**
+- Urban development monitoring
+- Disaster damage assessment
+- Property tax assessment
+- Solar panel installation planning