Mehmet Efe Akengin
MIT EECS Hewlett Foundation Undergraduate Research and Innovation Scholar
Cross-modal Visual Learning
Learning the features of different visual modalities like natural image, sketch and drawing and transferring features from one modality to another modality is an active area of research. The convolutional neural networks have shown a preliminary success in transferring the high level features to natural image modality. My work builds on the existing research done in MIT Vision Lab to create machine learning models that can learn the features of different modalities and transfer the high level features of one modality to another one.
I believe that SuperUROP is a fantastic opportunity to research in a more systematic way and be acknowledge for the research. I have been doing research in Vision Lab since Fall 2015 and I would like to continue working in this fantastic group. I hope to hone my skills in computer vision and deep learning and also work on a research project with more freedom. I am excited to be advancing the computer vision field.