This paper describes Asteroid, the PyTorch-based audio source separation toolkit for researchers. Inspired by the most successful neural source separation systems, it provides all neural building blocks required to build such a system. To improve reproducibility, Kaldi-style recipes on common audio source separation datasets are also provided. This paper describes the software architecture of Asteroid and its most important features. By showing experimental results obtained with Asteroid's recipes, we show that our implementations are at least on par with most results reported in reference papers. The toolkit is publicly available at github.com/mpariente/asteroid.
Car accidents are a major concern. Consequently, a lot of research is carried out on car user interfaces. For each such research, usually a special simulator or car is developed, algorithms and tools are redeveloped, and similar issues arise. We propose CarCoach, an educational car system, based on a generalized layered architecture. We present the system design, the intelligent modular architecture, its layers, including details of some of its relevant modules. Using the Chrysler 300M IT-Edition car as a platform, a prototype was implemented and initial experimentation was carried out and is reported. We demonstrate that CarCoach provides a flexible environment for car research and support of varied car applications.
In this article we offer a new approach to evaluating Organizational Memory (OM). Our proposed evaluation methodology, named KnowledgeEco, is based on an ontology for the domain of OM. Its key steps are: 1) mapping the OM in the evaluated organization onto the ontology concepts; 2) noting which entities from the ontology are missing in the OM; and 3) applying a series of rules that help assess the impact of the OM on organizational learning. This systematic evaluation thus helps to propose ways to improve the evaluated OM.We present three case studies that demonstrate the feasibility of KnowledgeEco for evaluating OM and for suggesting improvements. We also identify some weaknesses in the OMs common to the three organizations cited in the case studies. Finally, we discuss how the KnowledgeEco ontology-based methodology establishes utility and contributes to further research in the field of OM.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.