Karteek Alahari

Inria Grenoble - Rhône-Alpes - Thoth project-team
655 avenue de l'Europe, Montbonnot
38334 Saint Ismier Cedex
France (sending mail)

Email: karteek (dot) alahari @ inria (dot) fr
Tel: +33 4 56 52 71 31
Fax: +33 4 76 61 54 54
Karteek Alahari



I am a researcher (chargé de recherche Inria) in the Thoth research team (formerly known as LEAR), based at the Inria Grenoble - Rhône-Alpes center.

I was previously a postdoctoral fellow in the Inria WILLOW team at the Department of Computer Science, École Normale Supérieure (ENS), where I worked with Ivan Laptev, Jean Ponce, and Josef Sivic. I completed my Ph.D. in July 2010, under the supervision of Philip Torr. I am also an associate member of the Visual Geometry Group, University of Oxford and the WILLOW team at ENS.

My current research focuses on addressing the visual understanding problem in the context of large-scale datasets. In particular, I work on modelling tasks, such as pose estimation, recognition, tracking, segmentation, as optimization problems, learning parameters of the cost functions this induces with minimally-annotated data, and developing inference algorithms that scale-up to large data collections.




CV PUBLICATIONS SOFTWARE & DATASETS STUDENTS & POSTDOCS TEACHING
updated: May 2017 complete list, Google scholar, DBLP





NEWS (A complete list is here)
  • I will be an area chair (senior program committee member) for CVPR 2018.
  • New technical report on learning video object segmentation with visual memory is available.
  • Code for our CVPR 2017 paper on learning motion patterns is available.
  • Pre-print of our IJDAR paper, on scene text binarization, is available.
  • Code for our ICCV 2015 paper on tracking is available.
  • Promoted to senior member of IEEE (February 2017 list).
  • Former PhD student Yang Hua will start as a lecturer (asst. prof.) at Queen's Univ. Belfast, UK, in 2017.
  • Code for our ECCV 2016 paper on weakly-supervised segmentation methods in available.
  • Together with C. V. Jawahar, we are editing a Springer Advances in Computer Vision & Pattern Recog. volume on visual text interpretation.
  • We won the visual object tracking challenge (VOT-TIR track) held at ICCV 2015.
  • Received Google Faculty Research Award 2015 (see announcement).




    STUDENTS & POSTDOCS

    Postdocs: PhD students: Graduated: Masters: Former postdoc:




    TEACHING

    2016-'17: Introduction to Discrete Optimization (at École Centrale Paris)
    2016-'17: Understanding Big Visual Data (at ENSIMAG, in French)
    2015-'16: Introduction to Discrete Optimization (at École Centrale Paris)
    2014-'15: Multimedia databases (at ENSIMAG, in French)
    2013-'14: Multimedia databases (at ENSIMAG)





    RECENT NEWS from Inria (A complete list is here)
  • Inria has participated in the French national event, Fête de la Science. See the article in French
  • See Inria's white paper on Artificial Intelligence: Current challenges and Inria's engagement in English and French
  • An overview of Inria's research themes: in images, with French titles
  • Facebook has kindly offered us hardware as part of their GPU partnership program. See the announcement in English and French
  • Review of Inria's activities in 2015, in English, French
  • Inria's significant efforts for parity and equality are reviewed in a recent article (in French)





    PERSONAL






    *Photo courtesy: V. Peregrin.