Panoramic Video Salient Object Detection with Ambisonic Audio Guidance

TL;DR

This paper introduces a novel method for salient object detection in panoramic videos using ambisonic audio guidance, employing a multimodal fusion module with spherical positional encoding to enhance 3D spatial understanding.

Contribution

It presents a new multimodal fusion approach with pseudo-siamese ACF blocks and spherical positional encoding for panoramic video saliency detection.

Findings

Achieves state-of-the-art performance on ASOD60K dataset.

Effective audio-visual interaction through the proposed fusion module.

Enhanced spatial correspondence capturing in 3D panoramic videos.

Abstract

Video salient object detection (VSOD), as a fundamental computer vision problem, has been extensively discussed in the last decade. However, all existing works focus on addressing the VSOD problem in 2D scenarios. With the rapid development of VR devices, panoramic videos have been a promising alternative to 2D videos to provide immersive feelings of the real world. In this paper, we aim to tackle the video salient object detection problem for panoramic videos, with their corresponding ambisonic audios. A multimodal fusion module equipped with two pseudo-siamese audio-visual context fusion (ACF) blocks is proposed to effectively conduct audio-visual interaction. The ACF block equipped with spherical positional encoding enables the fusion in the 3D context to capture the spatial correspondence between pixels and sound sources from the equirectangular frames and ambisonic audios. Experimental results verify the effectiveness of our proposed components and demonstrate that our method achieves state-of-the-art performance on the ASOD60K dataset.