Skip to content

Latest commit

 

History

History
11 lines (7 loc) · 611 Bytes

README.md

File metadata and controls

11 lines (7 loc) · 611 Bytes

Models of the SQM paradigm with RNNs in TensorFlow

For the laboratory of Psychophysics, EPFL, Switzerland

This project aims at modelling visual grouping with the Sequential Metacontrast Paradigm (a vision paradigm).

  • Different data sets tested.
  • Different models of reconstructors (LSTM, PredNet, ...) and decoders tested (Dense, fully-convolutionnal, ...).
  • Different criteria, like minimum entropy, tested to model the "conscious percept" formation by the visual system and its discrete behavior.

The final project report is in sqm_models_Marie_Anselmet.pdf