From Large-scale Audio Tagging to Real-Time Explainable Emergency Vehicle Sirens Detection

2025-06-29

AI Summary

To address the lack of high-quality, large-scale datasets for emergency vehicle (EV) siren recognition and the excessive computational cost of existing models—hindering edge deployment—this paper proposes E2PANNs, a lightweight and interpretable audio event detection architecture. Built upon the PANNs framework, E2PANNs integrates lightweight convolutional modules, transfer learning, and optimization using an AudioSet EV subset, while enhancing interpretability via Guided Backpropagation and Score-CAM. Cross-domain training further improves robustness. Evaluated on multiple benchmarks, E2PANNs achieves state-of-the-art performance: 92.3% frame-level F1-score and a 27% reduction in event-level detection error. It enables real-time inference on embedded devices (e.g., Jetson Nano) with latency under 30 ms and reduces false positive rate by 41%, significantly improving safety and practicality in intelligent transportation and autonomous driving applications.

Technology Category

• Planning, Routing, and Scheduling: Activity and Plan Recognition
• Machine Learning: Learning on the Edge & Model Compression
• Computer Vision: Learning & Optimization for CV

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Security and Privacy: Privacy-enhancing technologies

Abstract

Accurate recognition of Emergency Vehicle (EV) sirens is critical for the integration of intelligent transportation systems, smart city monitoring systems, and autonomous driving technologies. Modern automatic solutions are limited by the lack of large scale, curated datasets and by the computational demands of state of the art sound event detection models. This work introduces E2PANNs (Efficient Emergency Pre trained Audio Neural Networks), a lightweight Convolutional Neural Network architecture derived from the PANNs framework, specifically optimized for binary EV siren detection. Leveraging our dedicated subset of AudioSet (AudioSet EV) we fine-tune and evaluate E2PANNs across multiple reference datasets and test its viability on embedded hardware. The experimental campaign includes ablation studies, cross-domain benchmarking, and real-time inference deployment on edge device. Interpretability analyses exploiting Guided Backpropagation and ScoreCAM algorithms provide insights into the model internal representations and validate its ability to capture distinct spectrotemporal patterns associated with different types of EV sirens. Real time performance is assessed through frame wise and event based detection metrics, as well as a detailed analysis of false positive activations. Results demonstrate that E2PANNs establish a new state of the art in this research domain, with high computational efficiency, and suitability for edge-based audio monitoring and safety-critical applications.

Problem

• Detect emergency vehicle sirens accurately in real-time
• Overcome lack of large-scale datasets for siren detection
• Optimize lightweight CNN for edge device deployment

Innovation

• Lightweight CNN for EV siren detection
• Fine-tuned on dedicated AudioSet EV subset
• Real-time edge deployment with interpretability