caghankoksal / vqa-med Goto Github PK

I'm Çağhan Köksal, MSc. student at Technical University of Munich, and a Machine Learning Engineer based in Münich. I'm passionate about working on complex problems and making them easier. I love learning about state of art research and leverage them to solve real world problems. I have worked on different tasks of machine learning such as Unsupervised (Video) Segmentation, Self-Supervised Learning, Image Generation and Manipulation, Multi-modal learning and Named Entity Recongition.

• Programming Languages: Python, C++,
• Tools & Technologies: Pytorch, Knet, React, Node.js, MongoDB

• Graph Neural Networks on Abdominal Organs - Explored capabilities Graph Neural Networks on organ mesh data. Use GNNs to predict Age, BMI, Sex, Height and weight of the patients.
• Explaining Medical Image Classifiers with Visual Question Answering Models - Research project on medical visual question answering problem. Explored capabilities of multi-modal Flamingo architecture and domain specific image and text encoders such as PubmedCLIP, PubmedBERT etc.
• Multimodal Emotion Recognition in Comics Worked on Emotion Recognition task on the Golden Age comics. Both text and image modalities are leveraged and fused to classify the emotion of the character.
• Face Generation In Golden Age Comics: Worked on Context-based Face Generation in Golden Age Comics (US Comics between the 1930s-1950s). The model predicts the masked face by giving consecutive comic book panels to the model with a randomly selected face masked at the last frame.

Connect with me on:

• LinkedIn
• Google Scholar
• CV

Feel free to reach out to me at here for any inquiries or collaboration opportunities.

Thanks for visiting my GitHub profile! Feel free to explore my repositories and don't hesitate to reach out. 😊