DeepAI AI Chat
Log In Sign Up

Differentiable Deep Clustering with Cluster Size Constraints

by   Aude Genevay, et al.

Clustering is a fundamental unsupervised learning approach. Many clustering algorithms – such as k-means – rely on the euclidean distance as a similarity measure, which is often not the most relevant metric for high dimensional data such as images. Learning a lower-dimensional embedding that can better reflect the geometry of the dataset is therefore instrumental for performance. We propose a new approach for this task where the embedding is performed by a differentiable model such as a deep neural network. By rewriting the k-means clustering algorithm as an optimal transport task, and adding an entropic regularization, we derive a fully differentiable loss function that can be minimized with respect to both the embedding parameters and the cluster parameters via stochastic gradient descent. We show that this new formulation generalizes a recently proposed state-of-the-art method based on soft-k-means by adding constraints on the cluster sizes. Empirical evaluations on image classification benchmarks suggest that compared to state-of-the-art methods, our optimal transport-based approach provide better unsupervised accuracy and does not require a pre-training phase.


page 1

page 2

page 3

page 4


Deep Distribution-preserving Incomplete Clustering with Optimal Transport

Clustering is a fundamental task in the computer vision and machine lear...

Deep Divergence-Based Approach to Clustering

A promising direction in deep learning research consists in learning rep...

Deep Shells: Unsupervised Shape Correspondence with Optimal Transport

We propose a novel unsupervised learning approach to 3D shape correspond...

Unsupervised Activity Segmentation by Joint Representation Learning and Online Clustering

We present a novel approach for unsupervised activity segmentation, whic...

Efficient Image Splicing Localization via Contrastive Feature Extraction

In this work, we propose a new data visualization and clustering techniq...

RECAL: Reuse of Established CNN classifer Apropos unsupervised Learning paradigm

Recently, clustering with deep network framework has attracted attention...

A Scalable Framework for Sparse Clustering Without Shrinkage

Clustering, a fundamental activity in unsupervised learning, is notoriou...