Jump to Content
Sergi Caelles

Sergi Caelles

I am a Research Scientist in the Media Understanding team at Google Research. Before, I was a PhD student at the Computer Vision Lab in Eidgenössische Technische Hochschule Zürich (ETH Zürich) under the supervision of Prof. Luc Van Gool and the guidance of Dr. Jordi Pont-Tuset. I received a Degree and M.Sc. in Telecommunications Engineering from the Universitat Politecnica de Catalunya (UPC). I have interned at Bell Labs with Prof. Sebastian Randel, Facebook with Prof. Fernando de la Torre, and Google Research with Prof. Vittorio Ferrari. My current interest is finding the right data for a specific application by leveraging signals from multiple vision models.
Authored Publications
Google Publications
Other Publications
Sort By
  • Title
  • Title, desc
  • Year
  • Year, desc
    Preview abstract We show how transformers can be used to vastly simplify neural video compression. Previous methods have been relying on an increasing number of architectural biases and priors, including motion prediction and warping operations, resulting in complex models. Instead, we independently map input frames to representations and use a transformer to model their dependencies, letting it predict the distribution of future representations given the past. The resulting video compression transformer outperforms previous methods on standard video compression data sets. Experiments on synthetic data show that our model learns to handle complex motion patterns such as panning, blurring and fading purely from data. Our approach is easy to implement, and we release code to facilitate future research. View details
    No Results Found