Gradient-Based Localization and Spatial Attention for Confidence Measure in Fine-Grained Recognition using Deep Neural Networks

Authors

  • Charles A. Kantor KLASS AI Research (AIR) MILA, Quebec Artificial Intelligence Institute Ecole Centrale Paris (ECP), Paris-Saclay University
  • Léonard Boussioux KLASS AI Research (AIR) MIT Ecole Centrale Paris (ECP), Paris-Saclay University
  • Brice Rauby Ecole Centrale Paris (ECP), Paris-Saclay University
  • Hugues Talbot Ecole Centrale Paris (ECP), Paris-Saclay University INRIA Paris, France

Keywords:

Fine-Grained Recognition, Deep Learning, Wildlife, Classification

Abstract

Both theoretical and practical problems in deep learning classification benefit from assessing uncertainty prediction. In addition, current state-of-the-art methods in this area are computationally expensive: for example,~\cite{loquercio2020general} is a general method for uncertainty estimation in deep learning that relies on Monte-Carlo sampling. We propose a new, efficient confidence measure later dubbed Over-MAP that utilizes a measure of overlap between structural attention mechanisms and segmentation methods. It does not rely on sampling or retraining. We show that the classification confidence increases with the degree of overlap. The associated confidence and identification tools are conceptually simple, efficient and of high practical interest as they allow for weeding out misleading examples in training data. Our measure is currently deployed in the real-world on widely used platforms to annotate large-scale data efficiently.

Downloads

Published

2021-05-18

How to Cite

Kantor, C. A., Boussioux, L., Rauby, B., & Talbot, H. (2021). Gradient-Based Localization and Spatial Attention for Confidence Measure in Fine-Grained Recognition using Deep Neural Networks. Proceedings of the AAAI Conference on Artificial Intelligence, 35(18), 15807-15808. Retrieved from https://ojs.aaai.org/index.php/AAAI/article/view/17900

Issue

Section

AAAI Student Abstract and Poster Program