VISER: Visual Self-regularization

Abstract

In this work, we propose the use of large set of unlabeled images as a source of regularization data for learning robust visual representation. Given a visual model trained by a labeled dataset in a supervised fashion, we augment our training samples by incorporating large number of unlabeled data and train a semi-supervised model. We demonstrate that our proposed learning approach leverages an abundance of unlabeled images and boosts the visual recognition performance which alleviates the need to rely on large labeled datasets for learning robust representation. To increment the number of image instances needed to learn robust visual models in our approach, each labeled image propagates its label to its nearest unlabeled image instances. These retrieved unlabeled images serve as local perturbations of each labeled image to perform Visual Self-Regularization (VISER). To retrieve such visual self regularizers, we compute the cosine similarity in a semantic space defined by the penultima

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyizadinia2018viser

Related papers

• Visual Representation Learning With Self-supervised Attention For Low-label High-data Regime (2022)5.49
• SUVR: A Search-based Approach To Unsupervised Visual Representation Learning (2023)3.58
• Learning Robust Visual-semantic Embeddings (2017)15.22
• Robust Cross-modal Representation Learning With Progressive Self-distillation (2022)12.33
• Self-supervising Fine-grained Region Similarities For Large-scale Image Localization (2020)15.85
• Improving Spatiotemporal Self-supervision By Deep Reinforcement Learning (2018)13.50
• Revisiting Contrastive Methods For Unsupervised Learning Of Visual Representations (2021)3.91
• Univse: Robust Visual Semantic Embeddings Via Structured Semantic Representations (2019)0.00