Acoular: accessible and reproducible sound localization in Python

Acoular：Python 中可访问且可重现的声音定位

• 批准号: 528753521
• 负责人: Professor Dr.-Ing. Ennes Sarradj
• 金额: --
• 依托单位: Fachgebiet Technische Akustik
• 依托单位国家: 德国
• 项目类别: Research Grants
• 资助国家: 德国
• 项目状态: 未结题
• 来源: https://gepris.dfg.de/gepris/projekt/528753521?language=en
• 关键词: Acoular; accessible; reproducible; sound; localization

Acoular is an open Python package for the analysis of multichannel signals from microphone arrays that provides algorithms for tasks such as sound source localization, sound field visualization, and advanced data filtering. It is widely used in the acoustic testing community and for related machine learning applications. The proposal aims to extend the user base to related fields such as robotics, ultrasound and hydro-acoustics and also to create a sustainable basis for future community-based development. The following work is foreseen: improve the current continuous integration process with a better test coverage of the code to address both quality assurance and reproducibility of results, refactor some sub-modules to enable new use cases and interfacing with other research software, develop more strict release and deprecation strategies as well contribution guidelines, extend the documentation with much more worked examples and benchmark data sets and extend training for potential users.

Acoular是一个开放的Python软件包，用于分析麦克风阵列的多通道信号，为声源定位、声场可视化和高级数据过滤等任务提供算法。它广泛用于声学测试社区和相关的机器学习应用程序。该提案旨在将用户基础扩大到机器人、超声波和水声学等相关领域，并为未来社区发展奠定可持续的基础。预计将开展以下工作：通过更好的代码测试覆盖率来改进当前的持续集成过程，以解决质量保证和结果的再现性，重构一些子模块以实现新的用例并与其他研究软件接口，制定更严格的发布和弃用策略以及贡献指南，用更多的工作实例和基准数据集扩展文档，并为潜在用户提供培训。