Latest Articles for Audio Processing

VampNet transforms music processing through innovative token modeling techniques.

A new model improves timing accuracy for lyrics in music applications.

New method improves speech recognition using only raw audio data.

New methods aim to hide speaker identities while maintaining speech clarity.

FlexiAST allows models to adapt to various audio patch sizes efficiently.

A new method addresses audio-visual segmentation challenges in noisy environments.

This study explores bias in audio models used for instrument recognition.

Research explores methods for identifying topics directly from audio recordings.

CMNet improves voice clarity by reducing echo in communication devices.

A new method to improve speech quality using energy-efficient networks.

MuReNN combines parametric and nonparametric models for improved audio analysis.

Introducing a new model for clearer speech in noisy environments.

A new method improves audio matching using images, enhancing realism in audio environments.

New techniques aim to improve audio quality by addressing packet loss.

New systems are designed to detect fake audio recordings with improved accuracy.

MoisesDB offers a detailed dataset for advanced music sound separation.

HierVST transforms voices seamlessly, enhancing audio quality without needing extensive data.

DAVIS offers a fresh way to tackle audio and visual sound separation.

New method uses ultrasonic sounds to confuse speech recognition systems without detection.

New methods enhance the accuracy of extracting singing melodies from mixed audio.

New methods aim to enhance audio captioning for better accuracy and efficiency.

New model improves speech clarity in noisy environments using innovative methods.

A study on Korean folk songs using modern analytical methods.

New model improves speech recognition in noisy environments by focusing on a single speaker.

New strategies to enhance training stability for music pitch classification.

A new method for accurate pitch detection in music and sound.

A new approach improves object segmentation in video using audio-visual integration techniques.

Meta-SELD enhances sound event localization in diverse environments.

A new system improves voice recognition in loud settings using advanced techniques.

Assessing the effectiveness of voice anonymization without losing natural sound.

New models enhance audio classification accuracy and resilience against noise and attacks.

A look at how XLS-R models improve audio quality assessment in online meetings.

New strategies improve speech clarity in noisy environments for better recognition.

New pruning methods enhance zero-shot multi-speaker text-to-speech model performance.

New methods improve keyword spotting using available reading speech data.

New single-step methods improve accuracy in formant tracking for speech sounds.

A new earbud design improves sound clarity using bone conduction technology.

A new lightweight model improves pitch estimation using self-supervised learning techniques.

New methods developed to identify fake songs amidst growing concerns.

Learn how technology helps categorize music genres efficiently.