Each project highlights my skills and attention to detail. I strive to deliver designs that meet and exceed expectations.

 

Projects:

Semantic Segmentation on CARLA Dataset (Thesis)

This project was completed as part of my Bachelor's thesis in Cognitive Science and Artificial Intelligence. The goal was to investigate semantic segmentation techniques for urban driving environments using the CARLA simulator. Semantic segmentation allows an AI system to classify every pixel in an image, helping autonomous vehicles better understand their surroundings.

As part of the project, I developed and evaluated three deep learning models. One model served as a baseline, while two alternative architectures were implemented and compared to determine which approach performed best on the selected dataset. Model performance was evaluated using segmentation metrics and visual inspection of the generated outputs.

This project introduced me to computer vision and deep learning on a larger scale. It gave me hands-on experience with training neural networks, comparing model architectures, and working with real-world AI challenges. Looking back, the project strengthened both my technical skills and my confidence in conducting independent research.

 

Raspberry Pi "Mushroom" Sleep Assistant

The Raspberry Pi Sleep Assistant was developed during my Master's programme in Human-Computer Interaction. The project focused on addressing a common challenge: reducing smartphone use before bedtime and encouraging healthier sleep habits.

Together with my team, I designed and developed an interactive device inspired by the Philips Somneo sleep system. The prototype took the form of a mushroom-shaped device that encouraged users to place their phone inside before going to sleep. Instead of using their smartphone, users were offered alternative activities such as listening to music, audiobooks, podcasts, or guided breathing exercises.

What I enjoyed most about this project was combining technology with an understanding of human behaviour. Unlike many of my previous AI-focused projects, this project required me to think about user needs, habits, and motivation. It showed me how technology can be used not only to solve technical problems but also to support positive behavioural change.

 

Pneumonia Detection Using Machine Learning

Technologies: Python, Scikit-learn, X-ray Imaging, Data Preprocessing

• Conducted a comparative analysis of four classifiers (KNN, SVM, MLP, and Logistic Regression) on

a dataset of 5,856 labeled chest X-ray images.

• Applied systematic data splitting (train/validation/test: 88%/11%/11%) and class balancing.

• Tuned hyperparameters for MLP including learning rate, batch size, hidden layers, and L2

regularization to optimize F1-score.

• Achieved robust classification performance, aiding in early and automated pneumonia diagnosis.

 

Deep Learning for Brain State Classification (MEG Data)

Technologies: PyTorch, MEG, Deep Learning, Signal Processing

• Developed deep learning models to classify cognitive tasks (rest, memory, math, motor) using

magnetoencephalography (MEG) signals.

• Implemented intra-subject and cross-subject classification, utilizing normalization and

downsampling for signal preprocessing.

• Explored model generalization across unseen subjects and adapted training with memory-efficient

data loaders.

• Investigated overfitting and improved accuracy through architecture tuning and alternative

modeling approaches.

 

Audio-based Classification with Neural Networks (this is a group project)

Technologies: PyTorch, Audio Signal Processing, CNNs

• Designed and trained a deep CNN model inspired by the "Look, Listen, and Learn" architecture for

audio classification.

• Replaced max-pooling with average pooling and added fully connected layers to improve model

expressiveness.

• Conducted hyperparameter tuning with learning rate schedules and dropout layers, reaching

70.7% validation accuracy.

• Used MFCCs for audio representation and achieved strong results despite small dataset size.