Introducing MERGE datasets to improve emotion classification in music.
― 6 min read
Computer Science - Sound
RSSThis study examines Mix-Training for keyword spotting in noisy speech conditions.
― 5 min read
A new method helps smaller models perform better using hints from larger models.
― 6 min read
Explore the updates in version 3 of the Divide and Remaster dataset.
― 6 min read
A comprehensive overview of datasets used in audio-language models and their importance.
― 9 min read
A reliable earbud-based system monitors breathing rates during various daily activities.
― 6 min read
Improving speech recognition systems for languages with limited online data.
― 5 min read
Combining sound and images for smarter recognition systems.
― 7 min read
A method to enhance audio deepfake detection through data augmentation.
― 5 min read
Beat-It generates synchronized dance movements to enhance choreography effortlessly.
― 5 min read
Researchers aim to create sounds that match silent videos, improving viewer experiences.
― 5 min read
This study addresses the issues with SLU systems and their ability to generalise.
― 6 min read
A self-supervised tool for estimating musical key signatures, reducing expert annotations.
― 5 min read
Diff-MST enhances music mixing by applying style transfer from reference tracks.
― 6 min read
A new model enhances communication for individuals with disabilities using speech recognition and Morse code.
― 5 min read
ElasticAST allows processing of variable length audio efficiently without losing important details.
― 5 min read
Analyzing singer identification methods amidst growing voice cloning concerns.
― 5 min read
A novel approach improves detection of mixed real and fake audio clips.
― 6 min read
Mamba shows promise against transformers in speech tasks, especially for long inputs.
― 4 min read
SingFlex offers innovative solutions for creating diverse singing voices efficiently.
― 5 min read
A study on the complexity of Irish traditional dance tunes using compression methods.
― 5 min read
RefinPaint enhances music creation by identifying and refining weak areas effectively.
― 6 min read
A new framework enhances speaker verification performance with limited data.
― 6 min read
Exploring new ways AI can collaborate with musicians through interpretation.
― 5 min read
CADE improves audio detection against evolving spoofing threats using continual learning techniques.
― 7 min read
A new method helps robots find fallen objects using sound.
― 5 min read
New voice command systems enhance drone control without the need for hands.
― 5 min read
New techniques allow for better emulation of guitar amplifiers and effects.
― 6 min read
A new framework enhances ASR performance using limited data and resources.
― 5 min read
A new method improves audio generation efficiency using innovative attention techniques.
― 5 min read
Discover how AI is transforming music generation with BandControlNet.
― 5 min read
A novel approach improves deepfake detection using audio-visual analysis.
― 5 min read
A look at the progress in speech recognition technologies and methods.
― 5 min read
A new method enhances stuttering detection by combining audio, video, and text data.
― 5 min read
A new method enhances sound creation for realistic 3D human models.
― 7 min read
This study reveals how speech can estimate breathing rates using advanced models.
― 5 min read
GraphMuse streamlines the analysis of symbolic music data with advanced machine learning techniques.
― 5 min read
Research presents new methods for evaluating speech recognition systems in Polish.
― 6 min read
A new dataset enhances machine speech for Mandarin, aiming for natural expression.
― 6 min read
A study on improving sound source localization by better using audio and visual information.
― 7 min read