Environmental sound classification is one of the important issues in the audio recognition field. Compared with structured sounds such as speech and music, the time-frequency structure of environmental sounds is more complicated. In order to learn time and frequency features from Log-Mel spectrogram more effectively, a temporal-frequency attention based convolutional neural network model (TFCNN) is proposed in this paper. Firstly, an experiment that is used as motivation in proposed method is designed to verify the effect of a specific frequency band in the spectrogram on model classification. Secondly, two new attention mechanisms, temporal attention mechanism and frequency attention mechanism, are proposed. These mechanisms can focus on key frequency bands and semantic related time frames on the spectrogram to reduce the influence of background noise and irrelevant frequency bands. Then, a feature information complementarity is formed by combining these mechanisms to more accurately capture the critical time-frequency features. In such a way, the representation ability of the network model can be greatly improved. Finally, experiments on two public data sets, UrbanSound 8 K and ESC-50, demonstrate the effectiveness of the proposed method.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC8566500 | PMC |
| http://dx.doi.org/10.1038/s41598-021-01045-4 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Conservation benefits of a large marine protected area network that spans multiple ecosystems.
Conserv Biol
January 2025
Marine Science Institute, University of California, Santa Barbara, Santa Barbara, California, USA.
Marine protected areas (MPAs) are widely implemented tools for long-term ocean conservation and resource management. Assessments of MPA performance have largely focused on specific ecosystems individually and have rarely evaluated performance across multiple ecosystems either in an individual MPA or across an MPA network. We evaluated the conservation performance of 59 MPAs in California's large MPA network, which encompasses 4 primary ecosystems (surf zone, kelp forest, shallow reef, deep reef) and 4 bioregions, and identified MPA attributes that best explain performance.
View Article and Find Full Text PDFUse and limitations of clinical data in the identification and classification of low molecular weight chemicals (LMWCs) as respiratory sensitizers: recommendations for improvement.
Crit Rev Toxicol
January 2025
Syngenta, Bracknell, United Kingdom.
While progress has been made in recent years, there are still no suitable and accepted , or models that can be used to accurately predict whether a chemical substance has the intrinsic property to cause immune-mediated chemical respiratory allergy, typically manifested as allergic asthma or allergic rhinitis which represents a severe health hazard. Regulatory authorities have relied primarily on clinical evidence (case reports, clinical databases, worker exposure studies) to classify substances as respiratory sensitizers, but this evidence can lack a proven immunological mechanism which is necessary to identify substances which can cause life-long sensitization and clinically relevant allergic symptoms in the respiratory tract in an exposed population (such respiratory allergens may be considered as "true" sensitizers, in analogy to the definition of skin sensitization, and in contrast to respiratory irritants). In light of this, the European Center for Ecotoxicology and Toxicology of Chemicals convened a Task Force to evaluate the types of clinical methods and data sources and the implications of relying on such data for regulatory decision making from a scientific perspective.
View Article and Find Full Text PDFThe recent ACHIEVE study (https://www.achievestudy.org/) demonstrated the substantial benefit of hearing aid use in those with mild-moderate hearing loss and at increased risk for cognitive decline.
View Article and Find Full Text PDFOne-quarter of freshwater fauna threatened with extinction.
Nature
January 2025
Tamar Valley National Landscape, Gunnislake, UK.
Freshwater ecosystems are highly biodiverse and important for livelihoods and economic development, but are under substantial stress. To date, comprehensive global assessments of extinction risk have not included any speciose groups primarily living in freshwaters. Consequently, data from predominantly terrestrial tetrapods are used to guide environmental policy and conservation prioritization, whereas recent proposals for target setting in freshwaters use abiotic factors.
View Article and Find Full Text PDFTuning into urban birdsong: enhancing nature connectedness with an AI-powered wearable.
Sci Rep
January 2025
School of Computer Science and Engineering, Southeast University, Nanjing, China.
The rapid urbanization has led to the loss of natural spaces and a subsequent disconnection between humans and nature, negatively affecting residents' well-being and environmental awareness. There is a a growing interest in leveraging technology to address this gap in Human-Computer Interaction. This article introduces GoChirp, an AI-powered wearable device for enhancing nature relatedness within urban landscapes.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!