Jitter-Adaptive Dictionary Learning - Application to Multi-Trial Neuroelectric Signals

Sebastian Hitziger (1) , Maureen Clerc (1) , Alexandre Gramfort (2) , Sandrine Saillet (3) , Christian G. Bénar (3) , Théodore Papadopoulo (1)

1 ATHENA - Computational Imaging of the Central Nervous System
2 LTCI - Laboratoire Traitement et Communication de l'Information
3 INS - Institut de Neurosciences des Systèmes

The simultaneous analysis of multiple recordings of neuronal electromagnetic activity is an important task requiring sophisticated and extremely noise robust techniques. A general goal is to find a representation of the similarities (e.g. repeating waveforms) as well as the differences (e.g. varying shape, latency, phase, or amplitude of waveforms) across the signals. Here, we present an extension to dictionary learning that explicitly accounts for small variations in latency and phase of atoms.

Domaines: Statistiques [stat]; Mathématiques [math]; Informatique [cs]

Sebastian Hitziger, Maureen Clerc, Alexandre Gramfort, Sandrine Saillet, Christian G. Bénar, et al.. Jitter-Adaptive Dictionary Learning - Application to Multi-Trial Neuroelectric Signals. International Conference on Learning Representations 2013, May 2013, Phoenix, Arizona, United States. 2013. ⟨hal-01094663⟩