Deep Convolutional K-Means Clustering

Anurag Goel (1) , Angshul Majumdar (1) , Emilie Chouzenoux (2) , Giovanni Chierchia (3, 4)

1 IIIT-Delhi - Indraprastha Institute of Information Technology [New Delhi]
2 OPIS - OPtimisation Imagerie et Santé
3 ESIEE Paris
4 LIGM - Laboratoire d'Informatique Gaspard-Monge

Conventional Convolutional Neural Network (CNN) based clustering formulations are based on the encoder-decoder based framework, where the clustering loss is incorporated after the encoder network. The problem with this approach is that it requires training an additional decoder network; this, in turn, means learning additional weights which can lead to over-fitting in data constrained scenarios. This work introduces a Deep Convolutional Transform Learning (DCTL) based clustering framework. The advantage of our proposed formulation is that we do not require learning the additional decoder network. Therefore our formulation is less prone to over-fitting. Comparison with state-of-the-art deep learning based clustering solutions on benchmark image datasets shows that our proposed method improves over the rest in challenging scenarios where there are many clusters with limited samples.

Mots clés: Convolutional Neural Network; K-means Clustering; Convolutional Transform Learning

Domaines: Traitement du signal et de l'image [eess.SP]

Anurag Goel, Angshul Majumdar, Emilie Chouzenoux, Giovanni Chierchia. Deep Convolutional K-Means Clustering. ICIP 2022 - 29th IEEE International Conference on Image Processing, Oct 2022, Bordeaux, France. ⟨hal-03723459⟩