Visualizing Celebrity Dynamics in Video Content: A Proposed Approach Using Face Recognition Timestamp Data

TL;DR

This work addresses the challenge of extracting and visualizing celebrity dynamics in video by integrating a distributed, multi-GPU face recognition pipeline with an interactive visualization platform. It introduces a scalable ONNX-based framework that produces timestamped $512$-dimensional embeddings, enabling precise temporal tracking of appearances across episodes, stored in Elasticsearch and indexed for rapid visualization. The visualization layer, powered by Plotly, offers a diverse suite of charts including co-appearance matrices and networks, heatmaps, and seasonal comparisons to reveal dominance, interactions, and temporal trends. Applied to real-world episodes, the approach yields actionable insights for entertainment analytics, content creation, and audience studies, while outlining future work on robustness and multi-modal integration.

Abstract

In an era dominated by video content, understanding its structure and dynamics has become increasingly important. This paper presents a hybrid framework that combines a distributed multi-GPU inference system with an interactive visualization platform for analyzing celebrity dynamics in video episodes. The inference framework efficiently processes large volumes of video data by leveraging optimized ONNX models, heterogeneous batch inference, and high-throughput parallelism, ensuring scalable generation of timestamped appearance records. These records are then transformed into a comprehensive suite of visualizations, including appearance frequency charts, duration analyses, pie charts, co-appearance matrices, network graphs, stacked area charts, seasonal comparisons, and heatmaps. Together, these visualizations provide multi-dimensional insights into video content, revealing patterns in celebrity prominence, screen-time distribution, temporal dynamics, co-appearance relationships, and intensity across episodes and seasons. The interactive nature of the system allows users to dynamically explore data, identify key moments, and uncover evolving relationships between individuals. By bridging distributed recognition with structured, visually-driven analytics, this work enables new possibilities for entertainment analytics, content creation strategies, and audience engagement studies.

Figures (12)

• Figure 1: Distributed Inference Framework Architecture
• Figure 2: Visualizer App Architecture
• Figure 3: Celebrity appearances per minute during the episode. The stacked bars represent the number of appearances for each celebrity at a given time. Episode: Ramazan Bey's Mansion, Season 2, Episode 20.
• Figure 4: Total number of appearances for each celebrity throughout the episode. Episode: Ramazan Bey's Mansion, Season 2, Episode 20.
• Figure 5: Total duration of appearances (in hours, minutes, and seconds) for each celebrity during the episode. Episode: Ramazan Bey's Mansion, Season 2, Episode 20.