Researcher

Juan Carlos Niebles

Also known as J. C. Niebles, J. Niebles, J.C. Niebles, Juan Niebles, Juan Carlos Niebles Duque

Juan Carlos Niebles is a computer vision researcher whose work spans video action recognition, multimodal models and AI agents. He earned engineering degrees at Universidad del Norte, the University of Illinois Urbana-Champaign and Princeton, and developed an unsupervised method for classifying and locating actions in video. He held academic and research posts at Universidad del Norte and Stanford, coauthored ActivityNet, and became Research Director at Salesforce in 2021. His Salesforce work included xGen-MM, xLAM and xGen-MM-Vid; Colombia named him to its Top 100 in AI in 2025. As of October 2026, he leads Samsung Research's North America AI Center as an executive vice president.

Timeline

  1. 2002–2011

    Engineering and video action words

    He earns an electronics engineering degree at Universidad del Norte in 2002. At the British Machine Vision Conference in 2006, he presents an unsupervised method that represents videos as spatial-temporal words and uses a probabilistic topic model to classify and locate human actions. He completes an M.Sc. in electrical and computer engineering at the University of Illinois at Urbana-Champaign in 2007 and a Ph.D. in electrical engineering at Princeton in 2011.

  2. 2011–2019

    Professor at Universidad del Norte

    From 2011 to 2019, he is an associate professor of electrical and electronic engineering at Universidad del Norte in Colombia.

  3. 2015–2021

    Stanford AI Lab and Toyota Center

    From 2015 to 2021, he is a senior research scientist at the Stanford AI Lab and associate research director at the Stanford-Toyota Center for AI Research. At CVPR 2015, he coauthors ActivityNet, a large-scale video benchmark for human activity understanding.

  4. 2021

    Research Director at Salesforce

    In 2021, he becomes Research Director at Salesforce and an adjunct professor of computer science at Stanford, where he co-directs the Stanford Vision and Learning Lab.

  5. 2024

    xGen-MM, xLAM, and video models

    In 2024, he coauthors xGen-MM (BLIP-3), an open framework for large multimodal models, and xLAM, a family of models optimized for function calling and AI agents. He also coauthors xGen-MM-Vid (BLIP-3-Video), a video language model that uses 32 visual tokens to represent a video.

  6. 2025

    Colombia's Top 100 in AI

    In March 2025, the Artificial Intelligence Observatory in Colombia names him among its Top 100 in AI.

  7. August 2026

    Head of North America AI Center

    In August 2026, he joins Samsung Research America as Head of the North America AI Center.

  8. As of October 2026

    Executive Vice President at Samsung

    He leads Samsung Research's North America AI Center as an executive vice president.

Career

  1. Since 2026SamsungExecutive Vice President and Head
  2. From 2021SalesforceResearch Director
  3. 2015–2021Stanford UniversitySenior Research Scientist, Adjunct Professor of Computer Science
  4. 2011–2019University of the NorthAssociate Professor of electrical and electronic engineering

Education

  1. 2011Princeton UniversityPh.D. in electrical engineering
  2. 2007University of Illinois Urbana-ChampaignM.Sc. in electrical and computer engineering
  3. 2002University of the NorthElectronics engineering degree

Sources and credits

Photo: We have no published photo of Juan Carlos Niebles yet. Send us a free-licensed photo.

  1. Juan Carlos Niebles Duque explorecourses.stanford.edu. Checked 21 October 2026.
  2. Juan Carlos Niebles, Author - Salesforce salesforce.com. Checked 21 October 2026.
  3. Unsupervised Learning of Human Action Categories Using Spatial-Temporal Words www-cs.stanford.edu. Checked 21 October 2026.
  4. ActivityNet: A Large-Scale Video Benchmark for Human Activity Understanding openaccess.thecvf.com. Checked 21 October 2026.
  5. [2408.08872] xGen-MM (BLIP-3): A Family of Open Large Multimodal Models arxiv.org. Checked 21 October 2026.
  6. xLAM: A Family of Large Action Models for AI Agents - Salesforce salesforce.com. Checked 21 October 2026.
  7. [2410.16267] xGen-MM-Vid (BLIP-3-Video): You Only Need 32 Tokens to Represent a Video Even in VLMs arxiv.org. Checked 21 October 2026.
  8. Celebrating Juan Carlos Niebles: Colombia's Top 100 in AI - Salesforce salesforce.com. Checked 21 October 2026.
  9. news | Juan Carlos Niebles niebles.net. Checked 21 October 2026.
  10. LATAM AI Summit · October 7, San Francisco | #SFTechWeek latamaisummit.vercel.app. Checked 21 October 2026.
  11. SAIF | Samsung Research research.samsung.com. Checked 21 October 2026.