ExposureEngine: Oriented Logo Detection and Sponsor Visibility Analytics in Sports Broadcasts

TL;DR

ExposureEngine addresses sponsor visibility analytics in sports broadcasts by adopting a rotation-aware Oriented Bounding Box detector to precisely localize logos under perspective distortions. It trains and evaluates on a new 1,103-frame soccer-focused dataset with 670 logos, achieving $mAP@0.5=0.859$, $precision=0.96$, and $recall=0.87$, and translates detections into exposure metrics via a polygon-based, logo-area framework. A LangGraph-based multi-agent layer enables natural-language querying, report generation, and clip sharing, making sponsor analytics auditable and accessible. The approach delivers more accurate on-screen coverage, supports real-time dashboards (≈$19.98$ FPS on GPU), and provides a practical blueprint for scalable sponsor valuation in broadcasts.

Abstract

Quantifying sponsor visibility in sports broadcasts is a critical marketing task traditionally hindered by manual, subjective, and unscalable analysis methods. While automated systems offer an alternative, their reliance on axis-aligned Horizontal Bounding Box (HBB) leads to inaccurate exposuremetrics when logos appear rotated or skewed due to dynamic camera angles and perspective distortions. This paper introduces ExposureEngine, an end-to-end system designed for accurate, rotation-aware sponsor visibility analytics in sports broadcasts, demonstrated in a soccer case study. Our approach predicts Oriented Bounding Box (OBB) to provide a geometrically precise fit to each logo regardless of the orientation on-screen. To train and evaluate our detector, we developed a new dataset comprising 1,103 frames from Swedish elite soccer, featuring 670 unique sponsor logos annotated with OBBs. Our model achieves a mean Average Precision (mAP@0.5) of 0.859, with a precision of 0.96 and recall of 0.87, demonstrating robust performance in localizing logos under diverse broadcast conditions. The system integrates these detections into an analytical pipeline that calculates precise visibility metrics, such as exposure duration and on-screen coverage. Furthermore, we incorporate a language-driven agentic layer, enabling users to generate reports, summaries, and media content through natural language queries. The complete system, including the dataset and the analytics dashboard, provides a comprehensive solution for auditable and interpretable sponsor measurement in sports media. An overview of the ExposureEngine is available online: https://youtu.be/tRw6OBISuW4 .

Figures (9)

• Figure 1: Example from a commercial logo detection solution using HBB thehive_ai. The HBB captures background regions around the rotated logo, leading to inaccurate size estimation.
• Figure 2: Class frequency distribution: number of classes per instance count category (log scale).
• Figure 3: ExposureEngine modular workflow from input video to analytical dashboard.
• Figure 4: Tightness Ratio vs. Orientation Necessity. Higher TR indicates a tighter fit of the OBB relative to its enclosing HBB.
• Figure 5: Loss components over training epochs for the YOLOv11-medium model, showing stable convergence for bounding box regression, classification, and distribution focal loss.