# A probabilistic constrained clustering for transfer learning and image category discovery Source: [A probabilistic constrained clustering for transfer learning and image category discovery]() Collection index: [llms.txt]() Data updated: 2026-01-02T12:57:55.997Z Authors: Yen-Chang Hsu, Zhaoyang Lv, Joel Schlosser, Phillip Odom, Zsolt Kira Published: 2018-06-28T16:49:19.000Z ## Abstract Neural network-based clustering has recently gained popularity, and in particular a constrained clustering formulation has been proposed to perform transfer learning and image category discovery using deep learning. The core idea is to formulate a clustering objective with pairwise constraints that can be used to train a deep clustering network; therefore the cluster assignments and their underlying feature representations are jointly optimized end-to-end. In this work, we provide a novel clustering formulation to address scalability issues of previous work in terms of optimizing deeper networks and larger amounts of categories. The proposed objective directly minimizes the negative log-likelihood of cluster assignment with respect to the pairwise constraints, has no hyper-parameters, and demonstrates improved scalability and performance on both supervised learning and unsupervised transfer learning. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Unsupervised character discovery on Omniglot cross-task from bg (eval)](): Unsupervised character discovery; Omniglot cross-task from bg (eval); Accuracy (100): 0.802. - [Image Category Discovery on ImageNet 30-class subsets (test)](): Image Category Discovery; ImageNet 30-class subsets (test); Accuracy: 74.4. ## Original sources and implementation [Original paper]() [PDF]()