Bipartite Graph Matching for Keyframe Summary Evaluation

TL;DR

This paper reviews graph-theoretic methods for evaluating keyframe video summaries, analyzes their behaviors through case studies, and recommends a specific greedy matching algorithm for improved evaluation accuracy.

Contribution

The paper provides a comparative analysis of frame matching methods for summary evaluation and advocates for a particular greedy algorithm based on their findings.

Findings

Different matching methods exhibit varied behaviors.

Case studies illustrate the strengths and weaknesses of each method.

The greedy matching algorithm by Kannappan et al. is recommended for better evaluation.

Abstract

A keyframe summary, or "static storyboard", is a collection of frames from a video designed to summarise its semantic content. Many algorithms have been proposed to extract such summaries automatically. How best to evaluate these outputs is an important but little-discussed question. We review the current methods for matching frames between two summaries in the formalism of graph theory. Our analysis revealed different behaviours of these methods, which we illustrate with a number of case studies. Based on the results, we recommend a greedy matching algorithm due to Kannappan et al.