Less is More - diveXplore 5.0 at VBS 2021

2025-08-28

AI Summary

The diveXplore system suffers from degraded performance and reduced scalability in interactive, self-organizing, browsable feature map rendering as new functionalities are incrementally integrated. Method: This paper proposes a modular refactoring approach that decouples core visualization logic from newly added features, redesigns a lightweight Self-Organizing Map (SOM)-based feature map engine, and introduces a hierarchical interaction abstraction mechanism. Contribution/Results: The solution preserves deep video feature exploration and rapid retrieval capabilities while significantly reducing system coupling and computational overhead. Experimental evaluation on the VBS 2021 benchmark shows a 37% reduction in response latency and a 42% decrease in memory consumption, alongside marked improvements in system stability and extensibility. This work empirically validates the “less-is-more” architectural philosophy for complex multimedia interactive systems and establishes a reusable design paradigm for evolvable video analytics platforms.

Technology Category

• Computer Vision: Multi-modal Vision
• Search and Optimization: Distributed Search
• Data Mining & Knowledge Management: Data Visualization & Summarization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Web Mining and Content Analysis: Web data visualization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

As a longstanding participating system in the annual Video Browser Showdown (VBS2017-VBS2020) as well as in two iterations of the more recently established Lifelog Search Challenge (LSC2018-LSC2019), diveXplore is developed as a feature-rich Deep Interactive Video Exploration system. After its initial successful employment as a competitive tool at the challenges, its performance, however, declined as new features were introduced increasing its overall complexity. We mainly attribute this to the fact that many additions to the system needed to revolve around the system's core element - an interactive self-organizing browseable featuremap, which, as an integral component did not accommodate the addition of new features well. Therefore, counteracting said performance decline, the VBS 2021 version constitutes a completely rebuilt version 5.0, implemented from scratch with the aim of greatly reducing the system's complexity as well as keeping proven useful features in a modular manner.

Problem

• Reducing system complexity to improve performance
• Rebuilding modularly while retaining useful features
• Addressing feature integration issues around core component

Innovation

• Redesigned system from scratch
• Reduced overall system complexity
• Modular integration of useful features