Page 5 - 《应用声学》2025年第3期
P. 5

第 44 卷 第 3 期                                                                       Vol. 44, No. 3
             2025 年 5 月                          Journal of Applied Acoustics                      May, 2025

             ⋄ 综述与评论 ⋄



                    无人机搜救场景下语声增强技术进展综述                                                             ∗





                                         雷 菁     1,2   杨飞然      1,2   杨 军    1,2†


                                      (1 中国科学院声学研究所噪声与音频声学实验室             北京   100190)
                                                (2 中国科学院大学      北京  100049)

                摘要：随着无人机技术的快速发展，无人机在各类场景中的应用日益广泛。在搜救任务中，无人机能够深入人
                力难以到达的区域，搜寻潜在的受困人员。无人机的应用不仅显著提升了搜救效率，还有效降低了救援人员面
                临的风险。相比图像信息，无人机搭载的传声器采集的声频信号在照明不足或视野受限的情况下，能够提供更
                加丰富和关键的线索。然而，由于环境噪声和无人机自噪声的干扰，采集到的信号通常信噪比极低，需要进行
                有效的增强处理。该文综述了无人机搜救场景下语声增强技术的最新研究进展，并重点探讨了无人机自噪声
                的特性、无人机场景中的语声增强技术及其面临的挑战。此外，该文还回顾了目前与无人机噪声相关的开源数
                据集，并展望了未来可能的发展方向。
                关键词：无人机搜救；语声增强；噪声消除
                中图法分类号: TN912.3           文献标识码: A          文章编号: 1000-310X(2025)03-0539-09
                DOI: 10.11684/j.issn.1000-310X.2025.03.001


                       A review of speech enhancement techniques in unmanned aerial

                                    vehicle-based search and rescue scenarios


                                       LEI Jing 1,2 , YANG Feiran 1,2  and YANG Jun 1,2

                        (1 Laboratory of Noise and Audio Acoustics, Institute of Acoustics, Chinese Academy of Sciences,
                                                    Beijing 100190, China)
                                   (2 University of Chinese Academy of Sciences, Beijing 100049, China)

                 Abstract: Drones are widely used in various scenarios with its rapid development. In search and rescue task,
                 drones can reach areas beyond human reach and search potential trapped individuals. The application of drones
                 significantly improves search and rescue efficiency and reduces the risks faced by rescuers. Compared to visual
                 information, the audio signals captured by microphones mounted on the drone can provide richer and more
                 critical clues, especially in conditions of poor lighting or limited visibility. However, due to the interference
                 of drone ego-noise, the captured signals often have a very low signal-to-noise ratio and need to be enhanced.
                 This review summarizes the latest research progress in speech enhancement techniques for unmanned aerial
                 vehicle(UAV)-based search and rescue scenarios, focusing on the characteristics of drone ego-noise, speech
                 enhancement methods in UAV scenarios and the challenges. In addition, this paper also reviews the currently
                 available open-source datasets related to drone noise and discusses potential future development directions.
                 Keywords: UAV search and rescue; Speech enhancement; Noise reduction


             2025-01-22 收稿; 2025-03-17 定稿
             国家自然科学基金项目 (62171438), 北京市自然科学基金项目 (4242013)
             ∗
             作者简介: 雷菁 (1999– ), 女, 河南信阳人, 博士研究生, 研究方向: 语声增强, 深度学习。
             † 通信作者 E-mail: jyang@mail.ioa.ac.cn
   1   2   3   4   5   6   7   8   9   10