Sharpen Focus: Learning with Attention Separability and Consistency
| Autor: | Bo Liu, Dimitris N. Metaxas, Ziyan Wu, Kuan-Chuan Peng, Rajat Vikram Singh, Srikrishna Karanam, Lezi Wang |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2018 |
| Předmět: |
FOS: Computer and information sciences
0209 industrial biotechnology Computer Science - Machine Learning Contextual image classification Computer science business.industry Computer Vision and Pattern Recognition (cs.CV) Computer Science - Computer Vision and Pattern Recognition 02 engineering and technology Pascal (programming language) Machine learning computer.software_genre Convolutional neural network Machine Learning (cs.LG) 020901 industrial engineering & automation Discriminative model 0202 electrical engineering electronic engineering information engineering 020201 artificial intelligence & image processing Artificial intelligence business computer computer.programming_language |
| Zdroj: | ICCV |
| Popis: | Recent developments in gradient-based attention modeling have seen attention maps emerge as a powerful tool for interpreting convolutional neural networks. Despite good localization for an individual class of interest, these techniques produce attention maps with substantially overlapping responses among different classes, leading to the problem of visual confusion and the need for discriminative attention. In this paper, we address this problem by means of a new framework that makes class-discriminative attention a principled part of the learning process. Our key innovations include new learning objectives for attention separability and cross-layer consistency, which result in improved attention discriminability and reduced visual confusion. Extensive experiments on image classification benchmarks show the effectiveness of our approach in terms of improved classification accuracy, including CIFAR-100 (+3.33%), Caltech-256 (+1.64%), ILSVRC2012 (+0.92%), CUB-200-2011 (+4.8%) and PASCAL VOC2012 (+5.73%). This paper is accepted to ICCV 2019. The supplementary material (appendix) can be found after the main paper |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|