New methods improve the quality of synthesized speech using self-supervised learning.
― 5 min read
Electrical Engineering and Systems Science - Audio and Speech Processing
RSSA new method enhances the transcription of rare keywords in business conversations.
― 6 min read
Federated Learning improves speech recognition while keeping user data private.
― 5 min read
MusicLDM transforms text into original music, offering fresh avenues for creativity.
― 7 min read
New methods enhance the accuracy of extracting singing melodies from mixed audio.
― 7 min read
New model improves speech clarity in noisy environments using innovative methods.
― 5 min read
A study on Korean folk songs using modern analytical methods.
― 8 min read
DiffDance creates detailed dance sequences that match music effectively.
― 5 min read
Examining fairness in singing voice transcription technology across genders.
― 8 min read
SeACo-Paraformer brings flexibility and accuracy to speech recognition technology.
― 5 min read
This study explores voice quality classification methods and their significance in communication.
― 4 min read
Learn how new algorithms improve noise cancellation techniques for various applications.
― 4 min read
AudioVMAF combines video metrics for improved audio quality assessment.
― 5 min read
A new method improves detection of fake audio using adaptive weight modification.
― 5 min read
Steganalysis helps detect hidden messages in multimedia, ensuring secure communication.
― 4 min read
A study on disentangling speaker identity from speech signals for improved processing.
― 5 min read
Transforming gestures for virtual agents with preserved meaning.
― 6 min read
Exploring how neural networks improve the accuracy of sound source localization.
― 6 min read
Researchers enhance automatic speech recognition for Punjabi using innovative self-training techniques.
― 5 min read
New model improves speech recognition in noisy environments by focusing on a single speaker.
― 4 min read
New methods aim to protect speech privacy in audio monitoring systems.
― 5 min read
A new dataset enhances speech synthesis by capturing emotional expression without relying on text.
― 5 min read
New strategies to enhance training stability for music pitch classification.
― 6 min read
Phoneme Hallucinator transforms voice conversion with limited data for clearer outputs.
― 5 min read
A new method creates realistic gestures from raw speech audio.
― 5 min read
Researchers develop Neural Latent Aligner to better interpret brain signals during speaking tasks.
― 6 min read
Enhancing hybrid ASR systems for bilingual speech using grapheme units.
― 5 min read
A new model improves speech and text alignment for better automatic recognition.
― 6 min read
Lip2Vec enhances visual speech recognition using fewer labeled data.
― 7 min read
New methods enhance accuracy and speed in speech recognition systems.
― 5 min read
O-1 improves speech recognition by optimizing self-training methods.
― 5 min read
A new method enhances ASR performance through text data integration.
― 6 min read
Text injection helps recognize personal information while maintaining privacy.
― 5 min read
Discover how new techniques are transforming sound event detection for various applications.
― 6 min read
Exploring nonlinear methods in audio for music production and speech analysis.
― 6 min read
A new method for accurate pitch detection in music and sound.
― 5 min read
Radio2Text uses mmWave signals for real-time speech recognition in noisy environments.
― 6 min read
A study examines the effectiveness of automated sound maskers in public spaces.
― 5 min read
Graph neural networks improve speaker recognition accuracy by analyzing voice sample relationships.
― 5 min read
A study evaluating emotion recognition in speech models across six languages.
― 5 min read