This method enhances music generation by separating emotional aspects into valence and arousal.
― 5 min read
Electrical Engineering and Systems Science - Audio and Speech Processing
RSSPiCoGen offers an innovative method for generating piano covers without paired data.
― 5 min read
Research focuses on identifying abusive speech in audio recordings across languages.
― 5 min read
A method to create audio that matches first-person viewpoint videos.
― 7 min read
A new system improves beat tracking across various musical genres.
― 5 min read
Study reveals listener views on AI-generated versus human music.
― 7 min read
A study on improving methods to detect lossy audio compression for better sound quality.
― 6 min read
This study examines how well LLMs understand and generate music.
― 5 min read
AI models enhance accuracy of speech-to-text conversions.
― 5 min read
Examining techniques to protect privacy while analyzing recorded conversations.
― 5 min read
An overview of MIDI music creation and its expressive potential.
― 5 min read
A new model that synchronizes chord annotations with music audio seamlessly.
― 5 min read
A new model integrates audio and visual data for speech recognition and translation.
― 6 min read
This study proposes a transparent way to assess music difficulty for educators.
― 6 min read
A new model enhances speech synthesis for various Chinese dialects.
― 5 min read
A new method improves piano cover creation, balancing quality and musical integrity.
― 4 min read
A framework that effectively identifies deepfake content through combined audio and visual analysis.
― 5 min read
A new benchmark to evaluate models analyzing music and language.
― 6 min read
A new framework improves classification in unseen audio-visual tasks.
― 6 min read
A new model enhances music generation using compound tokens and sequential decoding.
― 5 min read
A project reintroducing forgotten Korean court music using modern techniques.
― 6 min read
New methods enhance emotional expression in machine speech synthesis.
― 6 min read
A new method improves computer-generated music quality by separating melody and rhythm.
― 5 min read
This study examines how music and sounds evoke emotions together.
― 6 min read
New methods in AI music generation offer improved structure and diversity.
― 5 min read
New framework enhances speech recognition for diverse Arabic dialects.
― 4 min read
A system that creates unique drum rhythms based on written prompts for musicians.
― 4 min read
New methods improve speech recognition accuracy for diverse accents.
― 4 min read
A new method for judging how well audio pieces fit together in music.
― 5 min read
Methods to speed up speaker diarization without sacrificing accuracy.
― 6 min read
GRAFX offers an open-source solution for efficient audio processing with PyTorch.
― 4 min read
iDANSE enhances sound processing in acoustic sensor networks for better real-time applications.
― 4 min read
Improving binaural sound reproduction for better audio experiences in various devices.
― 7 min read
Wav2graph creates knowledge graphs from spoken language for improved AI understanding.
― 7 min read
Speech-MASSIVE aims to enhance spoken language understanding in various languages.
― 6 min read
Innovative techniques protect sensitive speech data while maintaining processing accuracy.
― 7 min read
Research on new models improves audio quality in film and television.
― 5 min read
New methods improve privacy while preserving speech content and emotions.
― 6 min read
Analyzing a child's sounds reveals crucial stages of language growth.
― 5 min read
New methods for better control of RNNs enhance audio effect simulations.
― 8 min read