My projects in deep learning and AI.
- Lockman hole, Optical identification of astrophysical X-ray sources with the help of neural networks, based on my paper.
- energAIser, A winning contribution to Siemens' Tech For Sustainability 2023 hackathon with time-series forecasting of energy consumption.
- Self Supervised Learning on Dark Matter halos, an ongoing project with application of self-supervised learning on (3D,2D) image and time series modalities and their alignment.
- RevolutionAIze, a contribution to TUM AI's hackathon with computer vision-based satellite image classification for deforestation detection.
- BlindVision, a contribution to Zeiss AG's Computer Vision hackathon with a neural network helping blind people during grocery shopping.
- Coursera ML, Coursera Machine Learning course by Andrew Ng with the implementation of basic algorithms and neural networks.
- Datacamp exam, practice exam for the certificate of "Data Scientist with Python" from Datacamp.
- Road segmentation model - find a road on an image from a car.
- Shapes 3D - convolutional autoencoder on the 3d shapes dataset.
- VAE synthetic data - Variational autoencoder on synthetic 2D data.
- Kitchenware Classification - classification of kitchenware items.
Books and lectures I find useful.
- Deep Learning, Ian Goodfellow and others
- Understanding Deep Learning, Simon J.D. Prince
- Little book of deep learning, François Fleuret
- Deep Learning Specialization, Andrew Ng and others (Coursera)
- Machine learning, Andrew Ng (Coursera, now unavailable)
- CS231n, Stanford Univ.