PeopleofSuperintelligence

Researcher

Karteek Alahari

Karteek Alahari is an Inria computer vision researcher whose work ranges from structured inference to video understanding and generative models. He completed a Ph.D. at Oxford Brookes in 2010 under Philip Torr, then moved from Inria's WILLOW team to LEAR in Grenoble in 2013. His 2010 work developed dynamic hybrid algorithms for MAP inference in discrete Markov random fields. He coauthored a memory-based video object segmentation system in 2017 and a multimodal transformer for natural-language video retrieval in 2020; Inria made him a research director in 2023. Since May 2024, he is Inria's Deputy Scientific Director for AI and a co-director of PEPR IA; he is an ELLIS Fellow, and in 2025 he coauthored Entropy Rectifying Guidance for diffusion and flow models.

Now at
Inria

Timeline

  1. 2010

    A computer vision PhD

    He completes a Ph.D. at Oxford Brookes University in July 2010 under Philip Torr, with a thesis titled Efficient Inference and Learning for Computer Vision Labelling Problems.

  2. 2010

    Dynamic hybrid algorithms

    With Pushmeet Kohli and Philip Torr, he coauthors a paper on dynamic hybrid algorithms for MAP inference in discrete Markov random fields in 2010.

  3. By 2013

    A postdoc in WILLOW

    He works as a postdoctoral fellow in Inria's WILLOW team at ENS's Department of Computer Science. In September 2013, he joins Inria's LEAR team in Grenoble as a non-permanent research scientist.

  4. October 2015

    A permanent Inria role

    He starts as a permanent researcher at Inria in October 2015.

  5. 2017–2020

    Memory for video understanding

    In 2017, he coauthors a two-stream neural network for video object segmentation with an explicit memory module. In 2020, he coauthors a multimodal transformer that jointly encodes video modalities for natural-language video retrieval.

  6. 2023

    Research director at Inria

    In June 2023, Inria promotes him to research director, effective in October 2023.

  7. 2024

    An ELLIS Fellowship

    In December 2024, he is selected as an ELLIS Fellow.

  8. Since 2024

    Deputy scientific director for AI

    Since May 2024, he is Inria's Deputy Scientific Director for AI and a co-director of the PEPR IA research programme. In 2025, he coauthors Entropy Rectifying Guidance, which changes attention at inference time to improve image quality, diversity and prompt consistency. He is also a Fellow in ELLIS's Multimodal Learning Systems program.

Career

  1. Since 2013InriaPostdoctoral Fellow, Non-permanent Research Scientist

Education

  1. 2010Oxford Brookes UniversityPh.D.

From the biography.

  1. People shaping AI work at a median of 3 organizations over a career: 3 for university starters and institute and lab starters, versus 2 for company startersIncludes Inria. Across 6,853 people shaping AI with a dated work role, we estimate a median of 3 parent organizations over a career (mean 3.6), comparing role groups, starting sectors and starting decades from the 1950s to 2020–2026.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Inria. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  3. Google has the highest combined h-index among AI organizationsIncludes Inria. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.
  4. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Inria. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.

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 Karteek Alahari yet. Send us a free-licensed photo.

  1. Karteek Alahari - Publications lear.inrialpes.fr. Checked 20 October 2026.
  2. Karteek Alahari - Homepage thoth.inrialpes.fr. Checked 20 October 2026.
  3. Karteek Alahari - News lear.inrialpes.fr. Checked 20 October 2026.
  4. Learning Video Object Segmentation with Visual Memory arxiv.org. Checked 20 October 2026.
  5. Multi-modal Transformer for Video Retrieval arxiv.org. Checked 20 October 2026.
  6. Entropy Rectifying Guidance for Diffusion and Flow Models papers.neurips.cc. Checked 20 October 2026.
  7. Karteek Alahari | European Laboratory for Learning and Intelligent Systems ellis.eu. Checked 20 October 2026.
  8. On the importance of a national AI strategy: interview with Fabien Le Voyer and Karteek Alahari inria.cl. Checked 20 October 2026.