Michael Gygli, Research Scientist at Google

Hi and welcome. My name is Michael Gygli and I am a Research Scientist at Google Zurich in the group of Prof. Vittorio Ferrari.

Previously I was the Head of AI at gifs.com, where I led the Machine Learning and Computer Vision efforts aimed at making video editing and gif creation intelligent and easy. Before that, I was a PhD student under the supervision of Prof. Luc Van Gool at the Computer Vision Laboratory of ETH Zurich. The research during my PhD was focused on video summarization, but I am also interested in deep learning, affective attributes, submodularity and data summarization in general.

Check the publications page or my Google Scholar profile for previous publications. I also provide code for some of my work on Github.

Slides for my CVPR tutorial talk "Video Summarization as Subset Selection": Click here


That's me (was, a few years ago)

Highlighted News

2018 February: I joined Google as a Research Scientist.
2017 December: We have released a new dataset for personalized highlight detection with data from more than 13000 users.
2017 July: The PathTrack paper was accepted to ICCV and the query-adaptive summarization paper to ACM Multimedia.
2017 June: Video and slides of my presentation "CNNs in Video analysis - An overview, biased to fast methods" at the ML meetup Zurich are now online.
2017 June: Code for our ICML paper on Deep Value Networks is now available on github.
2017 May: Our pre-print Ridiculously Fast Shot Boundary Detection with Fully Convolutional Neural Networks is now on arXiv.
2017 Jan: I joined gifs.com


Recent Papers

arXiv 2018

PHD-GIFs: Personalized Highlight Detection for Automatic GIF Creation
A. Garcia del Molino, M. Gygli

arXiv 2017

Interactive Video Object Segmentation in the Wild
A.Benard, M. Gygli

ICML 2017

Deep Value Networks Learn to Evaluate and Iteratively Refine Structured Outputs
M. Gygli, M. Norouzi, A. Angelova

arXiv 2017

Ridiculously Fast Shot Boundary Detection with Fully Convolutional Neural Networks
M. Gygli

Transactions on Multimedia 2017

AENet: Learning Deep Audio Features for Video Analysis
N. Takahashi, M. Gygli, L. Van Gool

ICCV 2017

PathTrack: Fast Trajectory Annotation with Path Supervision
S. Manen, M. Gygli, D. Dai, L. Van Gool