Toyota Jidosha Kabushiki Kaisha

Katholieke Universiteit Leuven

Wouter Van Gansbeke

Gabriel Othmezouri

Luc Van Gool

Marc Proesmans

Wim Abbeloos

Stamatios Georgoulis

Simon Vandenhende

A computer-implemented method for training a classifier (Φη), including: training a pretext model (ΦΘ) to learn a pretext task, so as to minimize a distance between an output of a source sample via the pretext model (ΦΘ) and an output of a corresponding transformed sample via the pretext model (ΦΘ), the transformed sample being a sample obtained by applying a transformation (T) to the source sample; S) determining a neighborhood (NXi) of samples (Xi) of a dataset (SD) in the embedding space; S) training the classifier (Φη) to predict respective estimated probabilities Φηj(Xi), j=1 . . . C, for a sample (Xi) to belong to respective clusters (Cj), by using a second training criterion which tends to: maximize a likelihood for a sample and its neighbors (Xj) of its neighborhood (Nxi) to belong to the same cluster; and force the samples to be distributed over several clusters.

Computer-implemented training method, classification method and system and computer-readable recording medium

Growing community of inventors

Dilbeek, Belgium

Wouter Van Gansbeke