Anmol Mishra

AI Researcher

Barcelona, Catalonia, Spain6 yrs 4 mos experience

AI ML PractitionerAI Enabled

Key Highlights

Expert in Generative AI and Deep Learning.
Contributed to open source audio analysis library Essentia.
Experience in high-impact AI projects at Samsung.

Stackforce AI infers this person is a Music Technology and AI specialist with strong software development skills.

Contact

Skills

Core Skills

Generative AiDeep LearningSoftware DevelopmentOpen Source DevelopmentArtificial Intelligence (ai)Machine Learning

Other Skills

Audio EngineeringC++PythonAmazon Web Services (AWS)Software DesignLinear AlgebraStatisticsSignalMathematical ModelingStatistical Data AnalysisData ModelingAndroid DevelopmentPyTorchJavaSQL

About

I am a Research Assistant at the Music Technology Group (MTG) and a master’s student at UPF Barcelona, where I contribute to the open source audio analysis library Essentia, and work with language models to learn expressive groove from Afro-Latin American drumming for expressive rhythms generation. I hold a Bachelor’s degree in Computer Science and Engineering from IIT Bombay and have professional experience as a Machine Learning Engineer at Samsung Headquarters in Seoul, where I contributed to the Computer Vision Group. Alongside my academic and technical work, I have been an active DJ, performing at music festivals and clubs in Seoul and Barcelona over the past two years.

Experience

6 yrs 4 mos

Total Experience

4 yrs

Average Tenure

2 yrs 3 mos

Current Experience

Music technology group

Research Assistant

Feb 2024 – Present · 2 yrs 3 mos · Barcelona, Catalonia, Spain · On-site

Expressive Groove Modeling for Afro-Latin Drumming
◦ Designed and developed a transformer based deep learning model to learn the distribution of timing and velocity deviations from a dataset of human drum performances in symbolic piano-roll representation.
◦ Devised a method to extend the approach for use with audio recordings and demonstrated its effectiveness with Uruguayan Candombe Rhythms.
◦ Developed a VST plugin for musicians to use the model in Digital Audio Workstations (DAW).
◦ Wrote a first-author paper on the method and presented during the Late Breaking Demo (LBD) Session at the ISMIR 2024 Conference in San Francisco.
Algorithms Development for Open Source MIR Libraries
◦ Contributed rhythm features, presented at ISMIR 2024, to the open source libraries Essentia and Carat.
◦ Developed optimized C++ implementations with Python bindings, ensuring compatibility with the APIs.
◦ Wrote tests for the features and developed a tutorial for users to integrate the features into their projects.
◦ Implemented a self-proposed plan to improve the Essentia documentation, restructuring the existing categories and migrating the user interface template to Read The Docs (RTD) format.

Generative AIDeep Learning