A new approach improves efficiency in multilingual ASR models by integrating adaptive masking techniques.
― 5 min read
Electrical Engineering and Systems Science - Audio and Speech Processing
RSSInvestigating deepfake audio to enhance transcription models for less common languages.
― 8 min read
New strategies enhance weak label learning by selecting relevant negative examples.
― 6 min read
A novel method to watermark audio created by diffusion models for ownership protection.
― 6 min read
New techniques enhance ASR systems for better long speech recognition.
― 5 min read
New techniques aim to boost the accuracy of voice-activated devices against attacks.
― 6 min read
DurIAN-E improves synthetic speech with enhanced expressiveness and natural flow.
― 4 min read
Discover how SER enhances human-machine interactions through emotion detection.
― 5 min read
A method to choose the best ASR model based on audio features.
― 5 min read
Learn how dereverberation boosts speech recognition in noisy environments.
― 4 min read
Coco-Nut offers diverse Japanese voice samples for advanced text-to-speech applications.
― 10 min read
This study presents an attention-based model for estimating room volumes from audio recordings.
― 5 min read
ASCA model enhances audio classification accuracy for small datasets.
― 5 min read
MyST aims to improve children's science learning through virtual tutoring.
― 5 min read
Study compares sound localization accuracy of four-channel and two-channel audio formats.
― 5 min read
A look at M2MeT 2.0 and its impact on meeting transcription.
― 5 min read
A new audio processing method enhances speaker anonymity while maintaining speech clarity.
― 5 min read
This study converts MRI tongue data into real speech audio.
― 4 min read
This study examines how model compression impacts speech recognition in noisy environments.
― 5 min read
Explore how Online Active Learning improves sound recognition efficiency.
― 6 min read
A new model improves understanding of speech and sounds simultaneously.
― 6 min read
A system that classifies client language in therapy sessions using multiple communication methods.
― 6 min read
New technology improves dysarthria detection and severity classification.
― 5 min read
New methods enhance early detection of voice problems using glottal source features.
― 5 min read
Enhancing speech models to better recognize and adapt to different accents.
― 4 min read
DCLS enhances audio classification performance by learning kernel positions during training.
― 5 min read
A new method enhances machine learning of audio-visual data.
― 5 min read
Introducing new models for better speech extraction in noisy environments.
― 5 min read
A new method enhances speech recognition efficiency using low-rank adaptation.
― 5 min read
Combining audio, video, and text for better mental health assessments.
― 5 min read
A look at advancements in speech recognition to boost speed and accuracy.
― 5 min read
Improving doctor-patient communication through advanced speech recognition technologies.
― 6 min read
Explore the privacy and security threats of voice-controlled technology.
― 4 min read
Synthia's Melody aids researchers in audio model testing against varied data.
― 5 min read
Research focuses on improving ASR systems for unsegmented audio.
― 4 min read
Research focuses on optimizing synthesizers for human vocalizations in various media.
― 5 min read
A new method improves speaker verification by managing session variability effectively.
― 6 min read
LLMs enhance accuracy and error correction in speech recognition systems.
― 5 min read
A new method enhances sound recognition and source location without labels.
― 5 min read
A new benchmark to improve ASR accuracy using language models.
― 6 min read