🔭 I'm currently working on:
Building an autoencoder for wireless communications as part of my MiSKS course at FTN. The idea is to model the whole tx/rx chain as a neural network and train it end-to-end instead of designing each block separately.
🤝 I'm looking to collaborate on:
Kaggle competitions, mostly anything involving computer vision or signal processing.
🤝 I'm looking for help with:
Getting better at deep learning. I know the basics of CNNs, RNNs and Transformers but still figuring out when and why to use each one.
🌱 I'm currently learning:
Deep learning architectures and how to actually apply them to real problems, not just run existing code.
💬 Ask me about:
My lip-reading classifier, learned image compression, signal processing, or any of my projects on GitHub.
⚡ Fun fact:
I built a classifier that tells apart Serbian and English just by watching lip movements. No audio at all, 65% accuracy against a 50% baseline.
4th year Telecommunications and Signal Processing student at the Faculty of Technical Sciences, University of Novi Sad.
