Patent attributes
Method(s) and apparatus are provided for operating and training an autoencoder. The autoencoder outputs a latent vector of an N-dimensional latent space for classifying input data. The latent vector includes a label vector y and a style vector z. The style vector z is regularised during training for effecting time invariance in the set of label vectors y associated with the input data. Method(s) and apparatus are further provided for controlling the optimisation of an autoencoder. The autoencoder outputting a latent vector of an N-dimensional latent space for classifying input data. The latent vector comprising a label vector y and a style vector z. The regularisation of the style vector z is controlled to increase or decrease the time invariance of the label vectors y. An autoencoder configured based on the above-mentioned trained autoencoder that regularised the style vector z for effecting time invariance in the set of label vectors y associated with the input data during classification.