Uncertainty measure of a mixture-model based pattern classifer

A classification device for determining an uncertainty measure of a mixture-model based parametric classifier, the classification device comprising processing circuitry, the processing circuitry being configured to cause the classification device to:

• obtain a short-term frequency representation of a multimedia signal, the short-term frequency representation defining an input sequence x;
• classify the input sequence x to belong to one class ωm* of at least two available classes ω1, ω2 using the parametric classifier, the parametric classifier having been trained with a training sequence y;
• and
• determine an uncertainty measure of the classified input sequence x based on a relation between posterior probabilities of the input sequence x and posterior probabilities of the training sequence y;
• and
• output at least the class ωm* and the determined uncertainty measure corresponding to the class ωm* for transmission.

A non-transitory computer storage medium storing a computer program for determining an uncertainty measure of a mixture-model based parametric classifier, the computer program comprising computer code which, when run on processing circuitry of a classification device, causes the classification device to:

• obtain a short-term frequency representation of a multimedia signal, the short-term frequency representation defining an input sequence x;
• classify the input sequence x to belong to one class ωm* of at least two available classes ω1, ω2 using the parametric classifier, the parametric classifier having been trained with a training sequence y;
• and
• determine an uncertainty measure of the classified input sequence x based on a relation between posterior probabilities of the input sequence x and posterior probabilities of the training sequence y;
• and
• output at least the class ωm* and the determined uncertainty measure corresponding to the class ωm* for transmission.