Initial commit

Author: v0lta

.flake8

@@ -0,0 +1,37 @@
+#########################
+# Flake8 Configuration  #
+# (.flake8)             #
+#########################
+[flake8]
+ignore =
+    S101
+    S301
+    S403
+    S404
+    S603
+    W503
+    E203
+    E402
+    DAR101
+    DAR201
+    N400
+exclude =
+    .tox,
+    .git,
+    __pycache__,
+    docs/source/conf.py,
+    build,
+    dist,
+    tests/fixtures/*,
+    *.pyc,
+    *.bib,
+    *.egg-info,
+    .cache,
+    .eggs,
+    data.
+max-line-length = 120
+max-complexity = 20
+import-order-style = pycharm
+application-import-names =
+    seleqt
+    tests

.github/workflows/test.yml

@@ -0,0 +1,23 @@
+name: Tests
+
+on: [ push, pull_request ]
+
+jobs:
+  lint:
+    name: Lint
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        python-version: [3.11.0]
+    steps:
+      - uses: actions/checkout@v2
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v2
+        with:
+          python-version: ${{ matrix.python-version }}
+      - name: Install dependencies
+        run: pip install nox
+      - name: Run flake8
+        run: nox -s lint
+      - name: Run mypy
+        run: nox -s typing

.gitignore

@@ -0,0 +1,174 @@
+ffhq_style_gan/
+ffhq_style_gan.zip
+
+.vscode/
+.pytest_cache/
+
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+classifier_weights.jpg
+input_opt.jpg
+mean_freq_difference.jpg
+real_fake_mean-log_fft2.jpg
+row_average_shifted_mean-log_fft2.jpg
+integrated_gradients_*.jpg
+out/

README.md

@@ -0,0 +1,75 @@
+# Exercise: Interpretable machine learning
+
+# Task 1: Input optimization.
+Open the `src/input_opt.py` file. The network `./data/weights.pth` contains network weights pre-trained on MNIST. Turn the network optimization problem around, and find an input that makes a particular output neuron extremely happy. In other words maximize,
+
+```math
+\max_\mathbf{x} y_i = f(\mathbf{x}, \theta) .
+```
+
+Use `torch.func.grad` to find the gradients of the network input $\mathbf{x}$.
+Start with a network input of shape `[1, 1, 28, 28]`. Compare a random initialization
+to starting from an array filled with ones, iteratively optimize it. Execute your script with `python src/input_opt.py`.
+
+# Task 2 Integrated Gradients (Optional):
+
+
+Reuse your MNIST digit recognition code. Implement IG as discussed in the lecture. Recall the equation
+
+```math
+\text{IntegratedGrads}_i(x) = (x_i - x_i') \cdot \frac{1}{m} \sum_{k=1}^m \frac{\partial F (x' + \frac{k}{m} \cdot (x - x'))}{\partial x_i}.
+```
+
+F partial xi denotes the gradients with respect to the input color-channels i.
+x prime denotes a baseline black image. And x symbolizes an input we are interested in.
+Finally, m denotes the number of summation steps from the black baseline image to the interesting input.
+
+Follow the todos in `./src/mnist_integrated.py` and then run `scripts/integrated_gradients.slurm`.
+
+
+
+# Task 3 Deepfake detection (Optional):
+In this exercise we will consider 128 by 128-pixel fake images from [StyleGAN](https://github.com/NVlabs/stylegan) and pictures of real people from the  [Flickr-Faces-HQ](https://github.com/NVlabs/ffhq-dataset) dataset.
+
+Flickr-Faces-HQ images depict real people, such as the person below:
+
+![real person](./figures/real.png)
+
+Generative adversarial networks allow the generation of fake images at scale. Does the picture below seem real? 
+
+![fake person](./figures/fake.png)
+
+How can we identify the fake? Given that modern neural networks can generate hundreds of fake images per second can we create a classifier to automate the process?
+
+### 3.1 Getting started:
+1. Move to the `data` folder in your terminal. Download [ffhq_style_gan.zip](https://drive.google.com/uc?id=1MOHKuEVqURfCKAN9dwp1o2tuR19OTQCF&export=download) on bender using the command
+   ```bash
+   gdown https://drive.google.com/uc?id=1MOHKuEVqURfCKAN9dwp1o2tuR19OTQCF
+   ```
+   If `gdown` is not installed, type `pip install gdown` and then try again.
+2. Type `export UNZIP_DISABLE_ZIPBOMB_DETECTION=TRUE` to make unzipping big archives possible.
+3. Extract the image pairs here by executing `unzip ffhq_style_gan.zip` in the terminal.
+
+The desired outcome is to have a folder called `ffhq_style_gan` in the project data-folder.
+
+
+### 3.2 Analyzing the data
+The `load_folder` function from the `util` module loads both real and fake data.
+Code to load the data is already present in the `deepfake_interpretation.py` file.
+
+Compute log-scaled frequency domain representations of samples from both sources via
+
+``` math
+\mathbf{F}_I =  \log_e (| \mathcal{F}_{2d}(\mathbf(I)) | + \epsilon ), \text{ with } \mathbf{I} \in \mathbb{R}^{h,w,c}, \epsilon \approx 0 .
+```
+
+Above `h`, `w` and `c` denote image height, width and columns. `Log` denotes the natural logarithm, and bars denote the absolute value. A small epsilon is added for numerical stability.
+
+Use the numpy functions `np.log`, `np.abs`, `np.fft.fft2`. By default, `fft2` transforms the last two axes. The last axis contains the color channels in this case. We are looking to transform the rows and columns.
+
+Plot mean spectra for real and fake images as well as their difference over the entire validation or test sets. For that complete the TODOs in `src/deepfake_interpretation.py` and run the script `scripts/train.slurm`.
+
+
+## 3.3 Training and interpreting a linear classifier
+Train a linear classifier consisting of a single `nn.Linear`-layer on the log-scaled Fourier coefficients using Torch. Plot the result. What do you see?
+