Gerald Friedland completed his Masters and Doctorate degrees in computer science from
Free University of Berlin in 2002 and 2006, respectively.[3] His PhD advisor was
Raúl Rojas. He then moved to the
International Computer Science Institute where he completed his a postdoc under
Nelson Morgan before continuing to be a research scientist and group leader there.[4] He then worked as a Principal Data Scientist at
Lawrence Livermore National Lab before co-founding Brainome, Inc.[5] He is a faculty fellow of the
Berkeley Institute for Data Science where he has been running a discussion group[6] since 2018, understanding the implications of using information theory as universal tool for modeling. This resulted in a book published in 2024.[7]
Career
Friedland is a
computer scientist specializing in the processing and analysis of
multimedia data and machine learning.[8] He is mostly known as the original author of the widely used "
Simple Interactive Object Extraction" image and video segmentation algorithm,[9][10][11][12][13][14][15][16] created as part of his PhD thesis,[17][18] and as the co-author of a textbook on Multimedia Computing.[19] He also led the initiative to create and release the YFCC100M corpus (see also:
List of datasets for machine learning research),[20][21][22] 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.[23][24][25] Friedland also frequently uncovers privacy risks in multimedia publishing practice[26][27][28][29][30][31][32][33] and heads the development of the teachingprivacy.org[34] 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-sourcespeech recognition board that allows the creation of cloudless voice interfaces[35] for
Internet of things devices.
^Gerald Friedland: "Adaptive Audio and Video Processing for Electronic Chalkboard Lectures", Lulu Publishing,
ISBN978-1430303886, December 2006. 2016 reprint:
ISBN978-3-659-97771-8, Lambert Publishing, November 2016.
^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
^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.
^Choi, Jaeyoung, Friedland, Gerald "Multimodal Location Estimation of Videos and Images",
Springer Publishing October 2014
^Nils Peters, Howard Lei, Gerald Friedland: "Room identification using acoustic features in a recording", US Patent US20140161270A1
^Gerald Friedland Bertrand Irissou: Method of facilitating construction of a voice dialog interface for an electronic system, US Patent Application US15382163.