A two-head loss function for deep Average-K classification - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Preprints, Working Papers, ... Year : 2023

A two-head loss function for deep Average-K classification

Abstract

Average-K classification is an alternative to top-K classification in which the number of labels returned varies with the ambiguity of the input image but must average to K over all the samples. A simple method to solve this task is to threshold the softmax output of a model trained with the cross-entropy loss. This approach is theoretically proven to be asymptotically consistent, but it is not guaranteed to be optimal for a finite set of samples. In this paper, we propose a new loss function based on a multi-label classification head in addition to the classical softmax. This second head is trained using pseudo-labels generated by thresholding the softmax head while guaranteeing that K classes are returned on average. We show that this approach allows the model to better capture ambiguities between classes and, as a result, to return more consistent sets of possible classes. Experiments on two datasets from the literature demonstrate that our approach outperforms the softmax baseline, as well as several other loss functions more generally designed for weakly supervised multi-label classification. The gains are larger the higher the uncertainty, especially for classes with few samples.
Fichier principal
Vignette du fichier
2303.18118.pdf (3.09 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04204318 , version 1 (17-10-2023)

Licence

Attribution

Identifiers

Cite

Camille Garcin, Maximilien Servajean, Joseph Salmon, Alexis Joly. A two-head loss function for deep Average-K classification. 2023. ⟨hal-04204318⟩
55 View
19 Download

Altmetric

Share

Gmail Facebook X LinkedIn More