Imagine Research News
Sound, Digested: MediaMined launches in Science 360 webcast, BBC World, Popular Mechanics, and more!
It's been a great week in the press: Imagine Research and the National Science Foundation partnered on the announcement of MediaMined.
We were honored to be featured in Science 360, BBC World, reports in major media outlets, and some of our favorite tech blogs.
Check out a behind the scenes look at our innovations in the Science 360 webcast, a 20-minute interview with CEO Jay LeBoeuf on the research and technology behind Imagine Research's MediaMined platform for indexing, understanding, and searching sound.
BBC World's Click discusses MediaMined, sound recognition, and audio search with Imagine Research. (starts at 6:45)Adding a set of ears to cloud computing.
Imagine Research announces MediaMined™, the world’s first sound object recognition web service. MediaMined identifies and labels sounds, making media files searchable and adding a set of ears to cloud computing.
This patent-pending technology allows companies to index their media based on the actual acoustic content - generating intelligent metadata and unique sound search signatures. This enables intelligent navigation with media files, improved search, discovery, and enables software and mobile devices workflows to finally hear their acoustic environment and input sounds.
Humans have an amazing ability to identify sounds. Within a fraction of a second, we can identify countless sound concepts: heavy metal music, a baby cry, James Earl Jones, a gun shot, or a dog bark. Using a proprietary, patent-pending audio signal processing and machine learning platform, along with massive datasets, Imagine Research allows computers to index, understand, and search sound.
Brewster Kahle, Digital Librarian and Founder of the Internet Archive, explains “Audio and video are some of the least discoverable assets on the Internet. If someone has not tagged the materials with the right keywords, they are likely to be lost. The Internet Archive is looking forward to using the automatic categorization technology of Imagine Research to help bring hundreds of thousands of audio and video files to a larger audience.”
MediaMined is in private beta with some of the world’s leading media companies, including music licensing, media creation and production, multimedia search, and digital asset management firms. Over system has analyzed nearly 2,000,000 unique media files – ranging from historical speeches to television sound tracks and sound effects.
Imagine Research, Inc is an intelligent audio technology company based in San Francisco.
Contact info at imagine-research.com for additional information.Intelligent Audio Systems: Foundations and Applications of Music Information Retrieval
Imagine Research is teaming up with researchers from Google, Princeton, and University of Victoria to teach at Stanford University’s Center for Computer Research in Music and Acoustics (CCRMA – pronounced “Karma”). This is the fourth year that Imagine Research’s CEO, Jay LeBoeuf, has led the course. The applied workshop teaches the underlying ideas, approaches, technologies, and practical design of intelligent audio systems using Music Information Retrieval (MIR) algorithms.
ASEE funds Imagine Research Fellowship
Imagine Research received was awarded Small Business Postdoctoral Research Diversity Fellowship, from the American Society for Engineering Education (ASEE). This program is designed to allow National Science Foundation Phase II Small Business Innovation Research (SBIR) awardees to accelerate their innovation research and to broaden participation from underrepresented groups.
Imagine Research welcomes Dr. Steve Tjoa as a Research Engineer to the team. Steve is a leading researcher in image and audio signal processing, machine learning, and source separation techniques. At the University of Maryland, Steve developed dictionary learning and sparse coding methods for music information retrieval tasks such as transcription, source separation, instrument recognition, and more. Steve’s image processing experience includes devising novel forensic methods to detect traces of compression in digital images based solely upon information intrinsic to the image and internships at Intelsat.
Mountain time represent! Welcome Derek.
Imagine Research now has mountain time zone represented. We're excited to welcome Derek Tingle as a research intern in machine learning and autotagging. As a NSF REU researcher Derek used machine learning techniques to decode neural signals. At Swarthmore College he developed a system to automatically label music. Derek will pursue a Master's degree from Stanford's CCRMA this fall.
Sound-object Recognition for Real-time or Offline Systems
Imagine Research was awarded a Phase II U.S. National Science Foundation’s Small Business Innovation Research (SBIR) program! This highly competitive $500,000 award was made to Imagine Research in recognition of the scientific and commercial potential of our MediaMined™ platform. MediaMined is the foundation of a broad, long-term, intelligent platform for audio analysis and processing.
In 2009, Imagine Research was awarded their first U.S. National Science Foundation Small Business Innovation Research (SBIR) Phase I grant in support of pioneering work in sound object recognition and search technology.
