Invariant Information Clustering for Unsupervised Image Classification and Segmentation

@article{Ji2019InvariantIC,
  title={Invariant Information Clustering for Unsupervised Image Classification and Segmentation},
  author={X. Ji and A. Vedaldi and Jo{\~a}o F. Henriques},
  journal={2019 IEEE/CVF International Conference on Computer Vision (ICCV)},
  year={2019},
  pages={9864-9873}
}
  • X. Ji, A. Vedaldi, João F. Henriques
  • Published 2019
  • Computer Science
  • 2019 IEEE/CVF International Conference on Computer Vision (ICCV)
  • We present a novel clustering objective that learns a neural network classifier from scratch, given only unlabelled data samples. The model discovers clusters that accurately match semantic classes, achieving state-of-the-art results in eight unsupervised clustering benchmarks spanning image classification and segmentation. These include STL10, an unsupervised variant of ImageNet, and CIFAR10, where we significantly beat the accuracy of our closest competitors by 6.6 and 9.5 absolute percentage… CONTINUE READING
    Self-labelling via simultaneous clustering and representation learning
    19
    GATCluster: Self-Supervised Gaussian-Attention Network for Image Clustering
    • 2020
    Learning To Classify Images Without Labels
    2
    Unsupervised Semantic Segmentation through Cross-Instance Representation Similarity
    Deep Transformation-Invariant Clustering
    Autoregressive Unsupervised Image Segmentation
    Mitigating Embedding and Class Assignment Mismatch in Unsupervised Image Classification

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 68 REFERENCES
    Deep Adaptive Image Clustering
    99
    Tagger: Deep Unsupervised Perceptual Grouping
    59
    Deep Clustering for Unsupervised Learning of Visual Features
    396
    Unsupervised Visual Representation Learning by Context Prediction
    813
    Deep Unsupervised Similarity Learning Using Partially Ordered Sets
    25
    Associative Deep Clustering: Training a Classification Network with No Labels
    31
    Joint Unsupervised Learning of Deep Representations and Image Clusters
    306
    Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks
    226
    Unsupervised Deep Embedding for Clustering Analysis
    683