ARD/ZDF award for Pelin Dogan Schönberger

Pelin Dogan Schönberger, postdoc with the Media Technology Center (MTC) of ETH Zurich, has received the first prize of the ARD/ZDF Förderpreis "Women + Media Technology" for her doctoral thesis titled "Automatic Alignment Methods for Visual and Textual Data with Narrative Content".

Dogan Schönberger's thesis focuses on aligning multi-modal data, specifically in visual and textual form, in order to learn and analyse correspondences between different manifestations of the same story. "An example can be aligning a movie to its written script, where data sequences to be aligned are a video sequence (composed of video shots) and a text sequence (composed of sentences), respectively", the researcher explains. She and her collaborators built upon recent advances in deep and recurrent neural networks, which provide efficient vectorial and contextual representations of the modalities to be aligned.

Such alignment methods can be used to automate meta data annotation, which is highly relevant for summarisation, archiving, and retrieval tasks. Considering the rapidly increasing amount of media content, including everything from books to videos to posts on the internet, automatic alignment is essential to alleviate the manual labour connected with these tasks.

"I am very happy to win this prize", Dogan Schönberger says. "I am grateful to my supervisor, Professor Markus Gross, who nominated me for this award, but also to the Department of Computer Science at ETH Zurich and to everyone who collaborated with me during my thesis." The researcher will continue exploring the intersection of machine learning, computer vision and natural language processing at Google Zurich starting in September.