%0 Journal Article %1 samuel2020longtail %A Samuel, Dvir %A Atzmon, Yuval %A Chechik, Gal %D 2020 %K feature-selection image-processing imbalanced %T Long-tail learning with attributes %U http://arxiv.org/abs/2004.02235 %X Learning to classify images with unbalanced class distributions is challenged by two effects: It is hard to learn tail classes that have few samples, and it is hard to adapt a single model to both richly-sampled and poorly-sampled classes. To address few-shot learning of tail classes, it is useful to fuse additional information in the form of semantic attributes and classify based on multi-modal information. Unfortunately, as we show below, unbalanced data leads to a "familiarity bias", where classifiers favor sample-rich classes. This bias and lack of calibrated predictions make it hard to fuse correctly information from multiple modalities like vision and attributes. Here we describe DRAGON, a novel modular architecture for long-tail learning designed to address these biases and fuse multi-modal information in face of unbalanced data. Our architecture is based on three classifiers: a vision expert, a semantic attribute expert that excels on the tail classes, and a debias-and-fuse module to combine their predictions. We present the first benchmark for long-tail learning with attributes and use it to evaluate DRAGON. DRAGON outperforms state-of-the-art long-tail learning models and Generalized Few-Shot-Learning with attributes (GFSL-a) models. DRAGON also obtains SoTA in some existing benchmarks for single-modality GFSL.

@article{samuel2020longtail, abstract = {Learning to classify images with unbalanced class distributions is challenged by two effects: It is hard to learn tail classes that have few samples, and it is hard to adapt a single model to both richly-sampled and poorly-sampled classes. To address few-shot learning of tail classes, it is useful to fuse additional information in the form of semantic attributes and classify based on multi-modal information. Unfortunately, as we show below, unbalanced data leads to a "familiarity bias", where classifiers favor sample-rich classes. This bias and lack of calibrated predictions make it hard to fuse correctly information from multiple modalities like vision and attributes. Here we describe DRAGON, a novel modular architecture for long-tail learning designed to address these biases and fuse multi-modal information in face of unbalanced data. Our architecture is based on three classifiers: a vision expert, a semantic attribute expert that excels on the tail classes, and a debias-and-fuse module to combine their predictions. We present the first benchmark for long-tail learning with attributes and use it to evaluate DRAGON. DRAGON outperforms state-of-the-art long-tail learning models and Generalized Few-Shot-Learning with attributes (GFSL-a) models. DRAGON also obtains SoTA in some existing benchmarks for single-modality GFSL.}, added-at = {2020-04-07T12:50:26.000+0200}, author = {Samuel, Dvir and Atzmon, Yuval and Chechik, Gal}, biburl = {https://www.bibsonomy.org/bibtex/25e86d888393990900800a93bcf69c240/kirk86}, description = {[2004.02235] Long-tail learning with attributes}, interhash = {06f83ac581dc5b09dd87d4910be498f3}, intrahash = {5e86d888393990900800a93bcf69c240}, keywords = {feature-selection image-processing imbalanced}, note = {cite arxiv:2004.02235}, timestamp = {2020-04-07T12:50:26.000+0200}, title = {Long-tail learning with attributes}, url = {http://arxiv.org/abs/2004.02235}, year = 2020 }