Time-frequency and time-scale analysis
Representations that reveal how signal structure evolves across time and frequency.
Adrien Meynard
I am an assistant professor (maître de conférences) in the Department of Physics at École Normale Supérieure de Lyon, where I work on signal processing with the Information, learning, and computation team.
Research interests
Signal processing
My research focuses on methods to analyse, model and interpret signals whose properties evolve over time.
Nonstationary signal modelling
Methods for describing changing dynamics, including amplitude and frequency modulation.
Applications
Signal processing for audio, biomedical engineering, physics and engineering science.
A research example
Separating the effect of motion from a sound
For a recording of a racing car, the Doppler effect and amplitude modulation can be modelled as deformations of an underlying stationary signal. The figures shown here display the scalograms of the original recording and the estimated stationary signal.
Illustration from the JEFAS algorithm.
JEFAS on GitHub
