---
title: "Uncertainty-Aware Few-Shot Image Classification"
authors:
  - "Zhizheng Zhang"
  - "Cuiling Lan"
  - "Wenjun Zeng"
  - "Zhibo Chen"
  - "Shih-Fu Chang"
doi: "10.24963/ijcai.2021/471"
url: "https://ijcai.org/proceedings/2021/471"
pdf: "https://ijcai.org/proceedings/2021/0471.pdf"
venue: "Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence"
year: 2021
pages: "3420-3426"
track: "Main Track"
keywords:
  - "Computer Vision: Recognition: Detection, Categorization, Indexing, Matching, Retrieval, Semantic Interpretation"
  - "Machine Learning: Cost-Sensitive Learning"
  - "Uncertainty in AI: Uncertainty Representations"
  - "Machine Learning: Classification"
---

# Uncertainty-Aware Few-Shot Image Classification

Zhizheng Zhang, Cuiling Lan, Wenjun Zeng, Zhibo Chen, Shih-Fu Chang

*Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence*, pages 3420-3426. DOI: [10.24963/ijcai.2021/471](https://doi.org/10.24963/ijcai.2021/471)

## Abstract

Few-shot image classification learns to recognize new categories from limited labelled data. Metric learning based approaches have been widely investigated, where a query sample is classified by finding the nearest prototype from the support set based on their feature similarities. A neural network has different uncertainties on its calculated similarities of different pairs. Understanding and modeling the uncertainty on the similarity could promote the exploitation of limited samples in few-shot optimization. In this work, we propose Uncertainty-Aware Few-Shot framework for image classification by modeling uncertainty of the similarities of query-support pairs and performing uncertainty-aware optimization. Particularly, we exploit such uncertainty by converting observed similarities to probabilistic representations and incorporate them to the loss for more effective optimization. In order to jointly consider the similarities between a query and the prototypes in a support set, a graph-based model is utilized to estimate the uncertainty of the pairs. Extensive experiments show our proposed method brings significant improvements on top of a strong baseline and achieves the state-of-the-art performance.