The Music and Audio Computing (MAC) Lab of the Academia Sinica in Taipei, part of its Research Center for Information Technology Innovation, was founded in Sept. 2011 by Dr. Yi-Hsuan Yang. The MAC Lab is dedicated to the development of multimedia systems that better understand what we hear and perceive in sounds and apply this understanding to enhance our interaction with sounds. Our goal is to carry out highly original and competitive research at the international level and transfer the technology to impact the daily life.

        Our current research interests include, but are not limited to: music emotion recognition, automatic tagging, music recommendation, multi-pitch estimation, lyrics processing, auditory scene analysis, audio event detection, singer/speaker identification, audio fingerprinting, human affect detection, and user interface design for mobile devices.

        We cordially welcome applicants who are interested in one of the following projects to work with us:

• “Hidden Patterns in Everyday Music Listening”
  Music listening is not a random activity; rather it depends on an interplay between some musical, personal, and situational factors. Being in a sad mood, some may listen to sad music, while others prefer happy tunes, for reasons. Our enthusiasm for a specific type of music is influenced by our peers. We tend to use music as “a reminder of valued past events.” Also our music listening is not randomly distributed in time; it follows some bursty yet regular pattern. This project aims at uncovering these hidden patterns so as to better understand human behavior and to create systems for music recommendation.

• “Musical Basis that Span Million Songs”
  Given the overwhelmingly diversity of music content, this project aims at discovering core musical elements from mining a million number of songs and associated metadata available from social websites. The goal is to decipher the language of music and provide a semantic description of music such as “This is a funky, dance pop song that is arousing not tender. It features male vocal, synthesizer, drum machine, a nice male vocal solo, and altered with effects, strong vocals. It is a song with a synthesized texture and that was recorded in studio that you might like listen to while at a party.”

        Or you may bring your own ideas. Qualification will be evaluated in terms of past research experience and the desire to excel in your career. In the attachment of the application letter please include at least your resume and purpose of doing research here. Shortlisted applicants would be invited for an on-site interview at Academia Sinica.