Exploring System Adaptations For Minimum Latency Real-Time Piano Transcription
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916942552498176 |
|---|---|
| author | Hu, Patricia Peter, Silvan David Schlüter, Jan Widmer, Gerhard |
| author_facet | Hu, Patricia Peter, Silvan David Schlüter, Jan Widmer, Gerhard |
| contents | Advances in neural network design and the availability of large-scale labeled datasets have driven major improvements in piano transcription. Existing approaches target either offline applications, with no restrictions on computational demands, or online transcription, with delays of 128-320 ms. However, most real-time musical applications require latencies below 30 ms. In this work, we investigate whether and how the current state-of-the-art online transcription model can be adapted for real-time piano transcription. Specifically, we eliminate all non-causal processing, and reduce computational load through shared computations across core model components and variations in model size. Additionally, we explore different pre- and postprocessing strategies, and related label encoding schemes, and discuss their suitability for real-time transcription. Evaluating the adaptions on the MAESTRO dataset, we find a drop in transcription accuracy due to strictly causal processing as well as a tradeoff between the preprocessing latency and prediction accuracy. We release our system as a baseline to support researchers in designing models towards minimum latency real-time transcription. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_07586 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Exploring System Adaptations For Minimum Latency Real-Time Piano Transcription Hu, Patricia Peter, Silvan David Schlüter, Jan Widmer, Gerhard Audio and Speech Processing Machine Learning Sound Advances in neural network design and the availability of large-scale labeled datasets have driven major improvements in piano transcription. Existing approaches target either offline applications, with no restrictions on computational demands, or online transcription, with delays of 128-320 ms. However, most real-time musical applications require latencies below 30 ms. In this work, we investigate whether and how the current state-of-the-art online transcription model can be adapted for real-time piano transcription. Specifically, we eliminate all non-causal processing, and reduce computational load through shared computations across core model components and variations in model size. Additionally, we explore different pre- and postprocessing strategies, and related label encoding schemes, and discuss their suitability for real-time transcription. Evaluating the adaptions on the MAESTRO dataset, we find a drop in transcription accuracy due to strictly causal processing as well as a tradeoff between the preprocessing latency and prediction accuracy. We release our system as a baseline to support researchers in designing models towards minimum latency real-time transcription. |
| title | Exploring System Adaptations For Minimum Latency Real-Time Piano Transcription |
| topic | Audio and Speech Processing Machine Learning Sound |
| url | https://arxiv.org/abs/2509.07586 |