Hyperbolic Audio-visual Zero-shot Learning

TL;DR

This paper introduces a hyperbolic space approach for audio-visual zero-shot learning, leveraging hyperbolic geometry to better model hierarchical data structures and improve classification accuracy on multiple datasets.

Contribution

It proposes a novel hyperbolic transformation with a new loss function and adaptive curvatures for cross-modality alignment in zero-shot learning.

Findings

Outperforms state-of-the-art methods on three datasets.

Achieves up to 7% improvement in harmonic mean.

Demonstrates the effectiveness of hyperbolic geometry in complex data structures.

Abstract

Audio-visual zero-shot learning aims to classify samples consisting of a pair of corresponding audio and video sequences from classes that are not present during training. An analysis of the audio-visual data reveals a large degree of hyperbolicity, indicating the potential benefit of using a hyperbolic transformation to achieve curvature-aware geometric learning, with the aim of exploring more complex hierarchical data structures for this task. The proposed approach employs a novel loss function that incorporates cross-modality alignment between video and audio features in the hyperbolic space. Additionally, we explore the use of multiple adaptive curvatures for hyperbolic projections. The experimental results on this very challenging task demonstrate that our proposed hyperbolic approach for zero-shot learning outperforms the SOTA method on three datasets: VGGSound-GZSL, UCF-GZSL, and ActivityNet-GZSL achieving a harmonic mean (HM) improvement of around 3.0%, 7.0%, and 5.3%, respectively.