Learning to Discover Novel Visual Categories via Deep Transfer Clustering
arXiv:1908.09884
Abstract
We consider the problem of discovering novel object categories in an image collection. While these images are unlabelled, we also assume prior knowledge of related but different image classes. We use such prior knowledge to reduce the ambiguity of clustering, and improve the quality of the newly discovered classes. Our contributions are twofold. The first contribution is to extend Deep Embedded Clustering to a transfer learning setting; we also improve the algorithm by introducing a representation bottleneck, temporal ensembling, and consistency. The second contribution is a method to estimate the number of classes in the unlabelled data. This also transfers knowledge from the known classes, using them as probes to diagnose different choices for the number of classes in the unlabelled subset. We thoroughly evaluate our method, substantially outperforming state-of-the-art techniques in a large number of benchmarks, including ImageNet, OmniGlot, CIFAR-100, CIFAR-10, and SVHN.
arXiv commentICCV 2019
References in corpus (7)
- Prototypical Networks for Few-shot Learning
- Deep Visual Domain Adaptation: A Survey
- Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results
- Towards K-means-friendly Spaces: Simultaneous Deep Learning and Clustering
- Learning to cluster in order to transfer across domains and tasks
- Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization
- Are Few-Shot Learning Benchmarks too Simple ? Solving them without Task Supervision at Test-Time
Cited by in corpus (5)
- EXPANSE: A Deep Continual / Progressive Learning System for Deep Transfer Learning
- A Review of Deep Transfer Learning and Recent Advancements
- Discovering New Intents with Deep Aligned Clustering
- Meta Discovery: Learning to Discover Novel Classes given Very Limited Data
- Augmented Data as an Auxiliary Plug-in Towards Categorization of Crowdsourced Heritage Data