Few-shot Partial Multi-view Learning

TL;DR

This paper introduces a new task called few-shot partial multi-view learning, addressing the challenge of classifying data with missing views and limited samples, and proposes a Gaussian dense-anchoring method to improve performance.

Contribution

The paper proposes a novel task and a unified Gaussian dense-anchoring method to handle view-missing and data scarcity issues simultaneously in few-shot multi-view learning.

Findings

The proposed method effectively alleviates the impact of view missing and data scarcity.

Extensive experiments demonstrate superior performance across multiple datasets.

The approach advances the capability of multi-view learning in low-data regimes.

Abstract

It is often the case that data are with multiple views in real-world applications. Fully exploring the information of each view is significant for making data more representative. However, due to various limitations and failures in data collection and pre-processing, it is inevitable for real data to suffer from view missing and data scarcity. The coexistence of these two issues makes it more challenging to achieve the pattern classification task. Currently, to our best knowledge, few appropriate methods can well-handle these two issues simultaneously. Aiming to draw more attention from the community to this challenge, we propose a new task in this paper, called few-shot partial multi-view learning, which focuses on overcoming the negative impact of the view-missing issue in the low-data regime. The challenges of this task are twofold: (i) it is difficult to overcome the impact of data scarcity under the interference of missing views; (ii) the limited number of data exacerbates information scarcity, thus making it harder to address the view-missing issue in turn. To address these challenges, we propose a new unified Gaussian dense-anchoring method. The unified dense anchors are learned for the limited partial multi-view data, thereby anchoring them into a unified dense representation space where the influence of data scarcity and view missing can be alleviated. We conduct extensive experiments to evaluate our method. The results on Cub-googlenet-doc2vec, Handwritten, Caltech102, Scene15, Animal, ORL, tieredImagenet, and Birds-200-2011 datasets validate its effectiveness.