Co-segmentation Inspired Attention Module for Video-based Computer Vision Tasks

TL;DR

This paper introduces COSAM, a co-segmentation inspired attention module that enhances CNN models for video tasks by automatically focusing on salient regions, improving performance and interpretability across multiple applications.

Contribution

The paper presents a novel, plug-and-play co-segmentation inspired attention module (COSAM) that improves video task performance without relying on pre-trained object models.

Findings

COSAM improves performance in video re-ID, captioning, and action classification.

COSAM generates interpretable attention maps highlighting task-specific regions.

The module is versatile and can be integrated into various CNN architectures.

Abstract

Video-based computer vision tasks can benefit from estimation of the salient regions and interactions between those regions. Traditionally, this has been done by identifying the object regions in the images by utilizing pre-trained models to perform object detection, object segmentation and/or object pose estimation. Although using pre-trained models is a viable approach, it has several limitations in the need for an exhaustive annotation of object categories, a possible domain gap between datasets, and a bias that is typically present in pre-trained models. In this work, we propose to utilize the common rationale that a sequence of video frames capture a set of common objects and interactions between them, thus a notion of co-segmentation between the video frame features may equip the model with the ability to automatically focus on task-specific salient regions and improve the underlying task's performance in an end-to-end manner. In this regard, we propose a generic module called ``Co-Segmentation inspired Attention Module'' (COSAM) that can be plugged in to any CNN model to promote the notion of co-segmentation based attention among a sequence of video frame features. We show the application of COSAM in three video-based tasks namely: 1) Video-based person re-ID, 2) Video captioning, & 3) Video action classification and demonstrate that COSAM is able to capture the task-specific salient regions in video frames, thus leading to notable performance improvements along with interpretable attention maps for a variety of video-based vision tasks, with possible application to other video-based vision tasks as well.