Sound Event Detection

Ask about this topic

Overview

Sound event detection (SED) aims to recognize acoustic events—such as speech, machinery, animal calls, or household sounds—and often locate them in time, sometimes also estimating their direction. Common approaches include CNNs, recurrent networks, and transformers trained with frame-level labels, alongside weakly supervised, semi-supervised, and self-supervised methods that reduce reliance on detailed annotation. Benchmarks such as DESED, AudioSet, and URBAN-SED support comparisons, while real-world bioacoustic, industrial, and domestic recordings test performance under overlapping sounds and changing environments. Key challenges are reliable temporal boundaries, limited labels and domain shift, and balancing accuracy with open-vocabulary capability and low-cost deployment.

Latest papers

CardsList

61 papers12 papers in the last four weeks, level with the four weeks before. 0.1% of all new papers.

Jul 28, 2025

Binaural Sound Event Localization and Detection based on HRTF Cues for Humanoid Robots

This paper introduces Binaural Sound Event Localization and Detection (BiSELD), a task that aims to jointly detect and localize multiple sound events using binaural audio, inspired by the spatial hearing mechanism of humans. To support this task, we present a synthetic benchmark dataset, called the Binaural Set, which simulates realistic auditory scenes using measured head-related transfer functions (HRTFs) and diverse sound events. To effectively address the BiSELD task, we propose a new input feature representation called the Binaural Time-Frequency Feature (BTFF), which encodes interaural time difference (ITD), interaural level difference (ILD), and high-frequency spectral cues (SC) from binaural signals. BTFF is composed of eight channels, including left and right mel-spectrograms, velocity-maps, SC-maps, and ITD-/ILD-maps, designed to cover different spatial cues across frequency bands and spatial axes. A CRNN-based model, BiSELDnet, is then developed to learn both spectro-temporal patterns and HRTF-based localization cues from BTFF. Experiments on the Binaural Set show that each BTFF sub-feature enhances task performance: V-map improves detection, ITD-/ILD-maps enable accurate horizontal localization, and SC-map captures vertical spatial cues. The final system achieves a SELD error of 0.110 with 87.1% F-score and 4.4° localization error, demonstrating the effectiveness of the proposed framework in mimicking human-like auditory perception.

Gyeong-Tae Lee, Hyeonuk Nam, Yong-Hwa Park

Korea Advanced Institute of Science and Technology, Daejeon, South Korea

Direction-of-Arrival EstimationSound Event Detection
AskPDF