MUS423 Research Seminars
The CCRMA Music 423 Research Seminar brings graduate students and supervising faculty together for planning and discussion of original research. Students and faculty meet either in small groups or individually, as appropriate for the research topics and interests of the participants. Research carried out is typically presented at the weekly CCRMA Colloquium (if it is of general interest to the CCRMA community) or at a Special DSP Seminar scheduled for that purpose. In either case, announcements appear on the CCRMA Home Page as Upcoming Events.
Recent DSP Seminars
Large-Scale Content-Based Matching of Audio and MIDI Data
Date:Tue, 01/13/2015 - 5:30pm - 7:30pmLocation:CCRMA Class Room [Knoll 217]Event Type:DSP Seminar"Large-Scale Content-Based Matching of Audio and MIDI Data"
FREEOpen to the PublicStructural Segmentation of Music using Posteriori Features
Date:Tue, 10/28/2014 - 6:00pm - 7:00pmLocation:CCRMA Classroom [Knoll 217]Event Type:DSP SeminarAbstract: Structural segmentation of music involves identifying boundaries between homogenous regions where the homogeneity involves one or more musical dimensions, and therefore depends on the musical genre. In this work, we address the segmentation of Hindustani instrumental concert recordings at the highest time-scale, that is, concert sections marked by prominent changes in rhythmic structure. Tempo features are effectively combined with energy and chroma features motivated by musicological knowledge and acoustic observations. Posterior probability features from unsupervised model fitting of the frame level acoustic features are shown to significantly improve robustness to local acoustic variations.
FREEOpen to the PublicBanging Like an 808? Developing a Physically-Informed, Circuit-Bendable, Digital Model of the Roland TR-808 Rhythm Composer
Date:Mon, 03/03/2014 - 5:30pm - 7:00pmLocation:CCRMA ClassroomEvent Type:DSP Seminar
Can a digital system provide satisfactory emulations of classic analog musical instruments and effects? Polarized reactions to the impending release (this month!) of Roland's TR-8 Rhythm Performer (a digitally modeled TR-808/909) have reignited this old debate. I've focused my research on creating physically-informed, circuit-bendable, digital models of analog drum machines. At this DSP Seminar, I'll present my methodological framework and analysis/models of one of the 808's most famous sound generators: the bass drum. My digital model, implemented in Cycling 74's Gen~, retain the salient features of the original analog drum machine and allow for accurate emulation of circuit-bending and common 808 mods. This work takes an engineering approach to hacking and circuit bending, clears up misconceptions about the 808's circuitry, and plugs a conspicuous drum-machine-shaped gap in virtual analog research.
FREEOpen to the PublicISSE: An Interactive Source Separation Editor, Part II
Date:Tue, 11/12/2013 - 5:30pm - 6:45pmLocation:CCRMA ClassroomEvent Type:DSP Seminar
Abstract: In the second talk of a two-part series, we will further discuss ISSE--an interactive source separation editor. As described earlier, ISSE is a new, open-source, freely available, cross-platform audio editing software tool which enables a user to perform single-channel source separation by painting on time-frequency visualizations of sound. It is useful for a variety of music- and audio-related tasks such as music remixing, audio denoising, and audio-based forensics.
FREEOpen to the PublicISSE: An Interactive Source Separation Editor, Part I
Date:Tue, 10/29/2013 - 5:30pm - 6:45pmLocation:CCRMA ClassroomEvent Type:DSP Seminar
Abstract: In the first talk of a two part-series, we will introduce ISSE--an interactive source separation editor. ISSE is a new, open-source, freely available, cross-platform audio editing software tool which enables a user to perform single-channel source separation by painting on time-frequency visualizations of sound. It is useful for a variety of music- and audio-related tasks such as music remixing, audio denoising, and audio-based forensics. In addition, we will discuss how to use the software, give demos, provide tips and tricks on achieving the high-quality separation, and outline limitations of the software. For more information, please see http://isse.sourceforge.net
FREEOpen to the PublicSource Separation Tutorial Mini-Series III: Extensions and Interpretations to Non-Negative Matrix Factorization
Date:Tue, 04/16/2013 - 5:30pm - 7:30pmLocation:CCRMA ClassroomEvent Type:DSP SeminarAbstract: Building off the last two lectures in the series, we will continue our discussion on non-negative matrix factorization techniques for source separation. We will talk about common extensions, additional interpretations, methods of evaluation, and if time permits, future directions of research. Following the lecture, we will get a chance to program and improve our basic separator from the second lecture. Please bring your laptops with Matlab and/or Octave installed and be ready to code!
Slides + Code @ https://ccrma.stanford.edu/~njb/teaching/sstutorial/FREEOpen to the PublicSource Separation Tutorial Mini-Series II: Introduction to Non-Negative Matrix Factorization
Date:Tue, 04/09/2013 - 5:30pm - 7:30pmLocation:CCRMA ClassroomEvent Type:DSP SeminarAbstract: For the second talk in this series, we will introduce the topic of non-negative matrix factorization for the purpose of single-channel source separation. NMF is one of the current most promising and effective class of approaches found for source separation and is a popular topic in several signal processing conferences and journals. Following the lecture, we will get a chance to program a basic source separator. Please bring your laptops with Matlab and/or Octave installed and be ready to code!
Slides + Code @ https://ccrma.stanford.edu/~njb/teaching/sstutorial/FREEOpen to the PublicSource Separation Tutorial Mini-Series I: Classical Speech Denoising and Enhancement
Date:Tue, 04/02/2013 - 5:30pm - 7:30pmLocation:CCRMA ClassroomEvent Type:DSP Seminar
Abstract: To start off a series of three tutorial-style dsp seminars on current single-channel source separation methods, the first talk will introduce the topic of classical methods used for speech enhancement. We will cover denoising methods such as spectral subtraction, wiener filters, and probabilistic estimators (if time permits). This will give an overview of the large range of literature on reducing background noise from speech.
FREEOpen to the PublicPOSTPONED! Source Separation Tutorial Mini-Series I: Classical Speech Denoising and Enhancement
Date:Tue, 02/19/2013 - 5:30pm - 7:00pmLocation:CCRMA Seminar Room [Knoll 315]Event Type:DSP Seminar
This event has been postponed. An announcement will be made when a new date is announced.FREEOpen to the PublicAudio Applications of Matrix Completion
Date:Tue, 11/27/2012 - 5:30pm - 7:00pmLocation:CCRMA Classroom [Knoll 217]Event Type:DSP Seminar"Audio Applications of Matrix Completion" --- Statistics PhD student Dennis Sun
FREEOpen to the Public
