Visual Recognition with Deep Nearest Centroids
We devise deep nearest centroids (DNC), a conceptually elegant yet surprisingly effective network for large-scale visual recognition, by revisiting Nearest Centroids, one of the most classic and simple classifiers. Current deep models learn the classifier in a fully parametric manner, ignoring the latent data structure and lacking simplicity and explainability. DNC instead conducts nonparametric, case-based reasoning; it utilizes sub-centroids of training samples to describe class distributions and clearly explains the classification as the proximity of test data and the class sub-centroids in the feature space. Due to the distance-based nature, the network output dimensionality is flexible, and all the learnable parameters are only for data embedding. That means all the knowledge learnt for ImageNet classification can be completely transferred for pixel recognition learning, under the "pre-training and fine-tuning" paradigm. Apart from its nested simplicity and intuitive decision-making mechanism, DNC can even possess ad-hoc explainability when the sub-centroids are selected as actual training images that humans can view and inspect. Compared with parametric counterparts, DNC performs better on image classification (CIFAR-10, ImageNet) and greatly boots pixel recognition (ADE20K, Cityscapes), with improved transparency and fewer learnable parameters, using various network architectures (ResNet, Swin) and segmentation models (FCN, DeepLabV3, Swin). We feel this work brings fundamental insights into related fields.
Code (1)
Tasks
Decision Makingimage-classificationImage ClassificationObject RecognitionMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
Improving the accuracy of nearest-neighbor classification using principled construction and stochastic sampling of training-set centroids
A conceptually simple way to classify images is to directly compare test-set data and training-set data. The accuracy of this approach is limited by the method of comparison used, and by the extent to which the training-…
ClassificationGeneral ClassificationLearning from One and Only One Shot
Humans can generalize from only a few examples and from little pretraining on similar tasks. Yet, machine learning (ML) typically requires large data to learn or pre-learn to transfer. Motivated by nativism and artificia…
Few-Shot LearningClassification with Nearest Disjoint Centroids
In this paper, we develop a new classification method based on nearest centroid, and it is called the nearest disjoint centroid classifier. Our method differs from the nearest centroid classifier in the following two asp…
Classificationfeature selectionMultilayer bootstrap networks
Multilayer bootstrap network builds a gradually narrowed multilayer nonlinear network from bottom up for unsupervised nonlinear dimensionality reduction. Each layer of the network is a nonparametric density estimator. It…
ClusteringDimensionality ReductionDistant Supervised Centroid Shift: A Simple and Efficient Approach to Visual Domain Adaptation
Conventional domain adaptation methods usually resort to deep neural networks or subspace learning to find invariant representations across domains. However, most deep learning methods highly rely on large-size source d…
Domain AdaptationDomain GeneralizationFace RecognitionPrivacy Preserving