Deep Unsupervised Learning of Visual Similarities
arXiv:1802.08562 · doi:10.1016/j.patcog.2018.01.036
Abstract
Exemplar learning of visual similarities in an unsupervised manner is a problem of paramount importance to Computer Vision. In this context, however, the recent breakthrough in deep learning could not yet unfold its full potential. With only a single positive sample, a great imbalance between one positive and many negatives, and unreliable relationships between most samples, training of Convolutional Neural networks is impaired. In this paper we use weak estimates of local similarities and propose a single optimization problem to extract batches of samples with mutually consistent relations. Conflicting relations are distributed over different batches and similar samples are grouped into compact groups. Learning visual similarities is then framed as a sequence of categorization tasks. The CNN then consolidates transitivity relations within and between groups and learns a single representation for all samples without the need for labels. The proposed unsupervised approach has shown competitive performance on detailed posture analysis and object classification.
arXiv commentarXiv admin note: text overlap with arXiv:1608.08792
References in corpus (5)
- DeepPose: Human Pose Estimation via Deep Neural Networks
- Beyond Short Snippets: Deep Networks for Video Classification
- Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICA
- Learning Fine-grained Image Similarity with Deep Ranking
- Shuffle and Learn: Unsupervised Learning using Temporal Order Verification