PeopleofSuperintelligence

Researcher

Henrik Stewénius

Also known as Henrik Stewenius

Henrik Stewénius is a computer vision researcher at Google in Switzerland who works on scalable image matching. After a 2005 doctorate at Lund University on Gröbner basis methods for minimal problems in computer vision, he published papers on three-view triangulation and direct relative orientation. At the University of Kentucky, he and David Nistér wrote "Scalable Recognition with a Vocabulary Tree" for CVPR in 2006; the scheme quantizes local image descriptors hierarchically in a tree, and the paper won the Longuet-Higgins Prize in 2016. With Steinar H. Gunderson and Julien Pilet, he presented "Size Matters" at ECCV 2012, a retrieval method that verifies the geometry of every candidate in a single pass through the index.

Timeline

  1. 2005

    A Lund doctorate on minimal problems

    He completes his doctoral thesis, "Gröbner Basis Methods for Minimal Problems in Computer Vision", at Lund University's Centre for Mathematical Sciences in 2005, supervised by Karl Åström. He defends it in April 2005, with William Triggs of INRIA Alpes as opponent. Its solvers cover relative pose for a generalised camera and for a camera with unknown focal length, optimal triangulation in three images, and hand-eye calibration.

  2. 2005–2006

    Triangulation and relative orientation

    With F. Schaffalitzky and D. Nistér, he presents "How Hard is Three-View Triangulation Really?" at ICCV 2005 in Beijing. In 2006, he coauthors "Recent developments on direct relative orientation" with C. Engels and Nistér in the ISPRS Journal of Photogrammetry and Remote Sensing.

  3. 2006

    The vocabulary tree

    At the University of Kentucky, he and David Nistér publish "Scalable Recognition with a Vocabulary Tree" at CVPR in 2006. The scheme quantizes local region descriptors hierarchically in a vocabulary tree, and a live demonstration recognizes CD covers from a database of 40,000 images.

  4. 2007–2008

    Hard-drive image search and MSER

    He coauthors "A Binning Scheme for Fast Hard Drive Based Image Search" with F. Fraundorfer and D. Nistér at CVPR 2007. At the 10th European Conference on Computer Vision in 2008, he and David Nistér present "Linear Time Maximally Stable Extremal Regions".

  5. December 2009

    A Google patent on located photos

    He is named as an inventor, with John Flynn, Ulrich Buddemeier, Hartmut Neven, Fernando Brucher and Hartwig Adam, on a Google patent application filed in December 2009. The invention matches a photo against a database of reference images that carry location information, and a match yields the photo's correct location and orientation.

  6. 2012

    Size Matters at ECCV

    With Steinar H. Gunderson and Julien Pilet, he publishes "Size Matters: Exhaustive Geometric Verification for Image Retrieval" at ECCV 2012; the method searches the index one document at a time and verifies the geometry of every candidate in a single pass. A Google patent application on image retrieval, filed in June 2012 with him as inventor, indexes the union of visual words across collections of near-duplicate images.

  7. 2016

    A Longuet-Higgins Prize

    As a Googler, he receives the Longuet-Higgins Prize at CVPR 2016 for the 2006 vocabulary tree paper with David Nistér. The retrospective award recognizes up to two CVPR papers that have made a significant impact on computer vision research.

  8. As of October 2026

    Computer vision at Google Switzerland

    He works on computer vision at Google in Switzerland. His research interests are scalable computer vision and image matching, with applications in Search By Image and Google Goggles.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Henrik Stewénius yet. Send us a free-licensed photo.

  1. Gröbner Basis Methods for Minimal Problems in Computer Vision | Lund University Publications lup.lub.lu.se. Checked 19 October 2026.
  2. How Hard is Three-View Triangulation Really? research.google. Checked 19 October 2026.
  3. Henrik Stewenius - Google Scholar scholar.google.com. Checked 19 October 2026.
  4. Scalable Recognition with a Vocabulary Tree | Sciweavers sciweavers.org. Checked 19 October 2026.
  5. CVPR 2016 & Research at Google research.google. Checked 19 October 2026.
  6. A Binning Scheme for Fast Hard Drive Based Image Search research.google. Checked 19 October 2026.
  7. Linear Time Maximally Stable Extremal Regions research.google. Checked 19 October 2026.
  8. US8189964B2 - Matching an approximately located query image against a reference image set - Google Patents patents.google.com. Checked 19 October 2026.
  9. Size Matters: Exhaustive Geometric Verification for Image Retrieval research.google. Checked 19 October 2026.
  10. US8923626B1 - Image retrieval - Google Patents patents.google.com. Checked 19 October 2026.
  11. Henrik Stewénius research.google. Checked 19 October 2026.