PhD Student Wins Best Paper Award at ICFHR 2018
This past August at the International Conference on Frontiers in Handwriting Recognition (ICFHR 2018), former CS PhD student Kenny Davila and CS Professor Richard Zanibbi received the Best Paper award. Their paper, "Visual Search Engine for Handwritten and Typeset Math in Lecture Videos and LaTeX Notes" is based on Kenny's PhD dissertation.
The paper presents the Tangent-V system, which can search handwriting in lecture videos using cropped images from the video, or images of typeset formulas as queries. The extraction of whiteboard contents from the video is entirely automatic, and captures what is on the whiteboard each second of the video. The system can be used to navigate through lecture videos by clicking on 'ink' in generated keyframes that capture everything written on the board. A demonstration of the system may be found online.
Competition for the award was stiff, as ICFHR is the leading international conference on handwriting recognition and applications. During the award presentation, the originality of the work was cited as the primary reason for selecting the paper. Kenny is currently a postdoctoral researcher at the CUBS lab at the University at Buffalo, continuing his work on document/video analysis and search.
(Shown in photo: Left to right: Srirangaraj Setlur (Organizing Chair), Kenny Davila, Christian Viard-Gaudin (Program Chair), and Richard Zanibbi )