Gerald Friedland

Last updated

Gerald Friedland is a German-American computer scientist and author specializing in multimedia computing, machine learning, and artificial intelligence. He is a principal scientist at Amazon Web Services and an adjunct professor at the electrical engineering and computer science department of the University of California, Berkeley. He focuses on AutoML and generative AI. His work has advanced large-scale multimedia analysis, privacy-aware AI, and explainable machine learning. [1] [2]

Contents

Education

Friedland completed his education in Germany, earning his Abitur in 1998. He received a Master of Science in computer science with a minor in linguistics from Freie Universität Berlin in 2002. His master’s thesis, "Towards a Generic Cross Platform Media Editor: An Editing Tool for E-Chalk," was recognized as the best computer science master’s thesis in German-speaking countries by the German Association for Computer Science. [3]

In 2006, Friedland earned his Ph.D. in Computer Science from Freie Universität Berlin, graduating summa cum laude. [4] His dissertation was entitled "Adaptive Audio and Video Processing for Electronic Chalkboard Lectures."

Career

Friedland began his career in academia as a Research Associate in the AI group at Freie Universität Berlin from 2002 to 2006 under Raúl Rojas. During this time, he developed the "Simple Interactive Object Extraction (SIOX)" algorithm, which is used in open-source tools such as GIMP and Blender, and conducted research on lecture webcasting technologies. [5]

From 2006 to 2016, Friedland worked full-time with the International Computer Science Institute (ICSI) in Berkeley, California. [6] He held various roles, from postdoctoral student to principal investigator to director of the audio and multimedia group. [7] As a principal data scientist at Lawrence Livermore National Laboratory (2016–2019), Friedland led a team addressing challenges in explainable AI. [8]

In 2014, he founded Audeme, a company developing cloud-independent speech recognition hardware. [9] [10] In 2019, he co-founded Brainome, Inc. which he joined full-time until 2022 as CTO, [7] leading the development of no-code machine learning solutions, leveraging the information-theory view of machine learning described in Information-Driven Machine Learning: Data Science as an Engineering Discipline. [11] [12]

Friedland served as director of conferences for ACM SIGMM (2017–2021), program co-chair for ACM Multimedia (2017), and associate editor for IEEE Multimedia Magazine and ACM Transactions on Multimedia Computing. [13] [14]

Research

Friedland is a computer scientist specializing in the processing and analysis of multimedia data and machine learning. [15] He is mostly known as the original author of the widely used "Simple Interactive Object Extraction" image and video segmentation algorithm, [16] [17] [18] [19] [20] [21] [22] [23] created as part of his PhD thesis, [24] [25] and as the co-author of a textbook on multimedia computing. [26] He also led the initiative to create and release the YFCC100M corpus (see also: List of datasets for machine learning research), [27] [28] [29] the largest freely available research corpus of consumer-produced videos and images. He co-founded the field of geolocation estimation for images and videos, sometimes also referred to as placing. [30] [31] [32] Friedland also frequently uncovers privacy risks in multimedia publishing practice [33] [34] [35] [36] [37] [38] [39] [40] and heads the development of the teachingprivacy.org [41] portal which provides educational materials for use in US high-schools as part of the AP Computer Science Principles and the Code.org initiative. Friedland is also the co-creator of MOVI, an open-source speech recognition board that allows the creation of cloudless voice interfaces [42] for Internet of things devices.

Awards

Publications

Friedland has authored six books, including:

He has also published over 100 peer-reviewed journal and conference articles on topics ranging from machine learning to multimedia computing. [15]

References

  1. ↑ "Gerald Friedland | EECS at UC Berkeley".
  2. ↑ "Gerald Friedland".
  3. ↑ "Mitteilungen der Gesellschaft für Informatik · 159. Folge (Fortsetzung)". Informatik-Spektrum (in German). 26 (1): 60–69. 2003-01-01. doi:10.1007/s002870200282. ISSN   1432-122X.
  4. ↑ "Refubium - Suche".
  5. ↑ Friedland, G.; Jantz, K.; Rojas, R. (2005). "SIOX: Simple Interactive Object Extraction in Still Images": 253–260. doi:10.1109/ism.2005.106.{{cite journal}}: Cite journal requires |journal= (help)
  6. ↑ "Gerald Friedland | ICSI". www.icsi.berkeley.edu. Retrieved 2024-12-19.
  7. 1 2 "UC Berkeley-Affiliated ICSI Launches Two New Groups for Audio/Multimedia and Research Initiatives in Computer Science". Yahoo Finance .
  8. ↑ "Efficient Saliency Maps for Explainable AI". Cornell University .
  9. ↑ "An interview with Bertrand and Gerald of Audeme | The Amp Hour Electronics Podcast". theamphour.com. 2015-07-16. Retrieved 2025-02-13.
  10. ↑ "Brainome launches product to optimize machine learning development process". ZDNet .
  11. ↑ Friedland, Gerald "Information-Driven Machine Learning: Data Science as an Engineering Discipline", Springer-Nature, January 2024.
  12. ↑ Woodie, Alex (2020-11-04). "Brainome Right-Sizes Your Data Before ML Training". BigDATAwire. Retrieved 2024-12-19.
  13. ↑ "New SIGMM Leadership Announced | ACM SIGMM - the Special Interest Group on Multimedia". www.sigmm.org. Retrieved 2024-12-19.
  14. ↑ "Gerald Friedland - Home". Author DO Series. Retrieved 2024-12-19.
  15. 1 2 Google Scholar list of publications: https://scholar.google.com/citations?user=iBl-QgEAAAAJ
  16. ↑ "Algorithm - What are the standard techniques for removing a segmentation (Such as a human or bird) from a video?".
  17. ↑ "SIOX".
  18. ↑ "Using GIMP's Foreground select tool". 31 August 2013.
  19. ↑ "Paintshopprotutorials.co.uk".
  20. ↑ "Kutout - an application for cutting out images | Hook - Labs". Archived from the original on 2017-07-24. Retrieved 2017-07-16.
  21. ↑ "Fiji plugin based on the SIOX project to segment color images: Fiji/Siox_Segmentation". GitHub . June 2019.
  22. ↑ "SIOX: Simple Interactive Object Extraction".
  23. ↑ Shoou Jiah Yiu, Gerald Friedland: "Method and system for identifying objects in images" US Patent Application US20170132469A1
  24. ↑ Gerald Friedland: "Adaptive Audio- und Videoverarbeitung für elektronische Kreidetafelvorlesungen", Freie Universitaet Berlin, October 2006. http://www.diss.fu-berlin.de/diss/receive/FUDISS_thesis_000000002354
  25. ↑ Gerald Friedland: "Adaptive Audio and Video Processing for Electronic Chalkboard Lectures", Lulu Publishing, ISBN   978-1430303886, December 2006. 2016 reprint: ISBN   978-3-659-97771-8, Lambert Publishing, November 2016.
  26. ↑ Friedland, Gerald and Jain, Ramesh "Multimedia Computing", Cambridge University Press, October 2014.
  27. ↑ Bart Thomee, David A. Shamma, Gerald Friedland, Benjamin Elizalde, Karl Ni, Douglas Poland, Damian Borth, Li-Jia Li. "YFCC100M: The New Data in Multimedia Research". Communications of the ACM, Vol. 59 No. 2, Pages 64-73
  28. ↑ YFCC100M: YFCC100M
  29. ↑ The Multimedia Commons
  30. ↑ Gerald Friedland, Oriol Vinyals, and Trevor Darrell: "Multimodal Location Estimation", in Proceedings of the ACM International Conference on Multimedia (ACM Multimedia 2010), Florence, Italy, October 2010, pp. 1245-1251.
  31. ↑ Choi, Jaeyoung, Friedland, Gerald "Multimodal Location Estimation of Videos and Images", Springer Publishing October 2014
  32. ↑ Nils Peters, Howard Lei, Gerald Friedland: "Room identification using acoustic features in a recording", US Patent US20140161270A1
  33. ↑ Web Photos That Reveal Secrets, Like Where you Live (New York Times, Aug 11, 2010)
  34. ↑ Tips to Turn Off Geo-Tagging on Your Cell Phone (ABC News, Aug 20, 2010)
  35. ↑ Could you fall victim to crime simply by geotagging location info to your photos? (Digital Trends, Jul 22, 2013)
  36. ↑ Ways to Avoid Email Tracking (New York Times, Dec 25, 2014)
  37. ↑ BodyWorn, the police-worn camera that aims to reduce crime (Fox News, May 19, 2015)
  38. ↑ Paris ISIS Attacks: Tech Industry Says 'Anti-Terror' Back Doors Would Make US Less Safe (International Business Times, Nov 18, 2015)
  39. ↑ Why our Crazy Smart AI still sucks at Transcribing our Speech (Wired Magazine, Apr 8, 2016)
  40. ↑ Transcribing Audio Sucks—So Make Machines Like Trint Do It (Wired Magazine, Apr 26, 2017)
  41. ↑ "Teaching Privacy".
  42. ↑ Gerald Friedland Bertrand Irissou: Method of facilitating construction of a voice dialog interface for an electronic system, US Patent Application US15382163.
This page is based on this Wikipedia article
Text is available under the CC BY-SA 4.0 license; additional terms may apply.
Images, videos and audio are available under their respective licenses.