From Weak to Strong Sound Event Labels using Adaptive Change-Point Detection and Active Learning
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909295505833984 |
|---|---|
| author | Martinsson, John Mogren, Olof Sandsten, Maria Virtanen, Tuomas |
| author_facet | Martinsson, John Mogren, Olof Sandsten, Maria Virtanen, Tuomas |
| contents | We propose an adaptive change point detection method (A-CPD) for machine guided weak label annotation of audio recording segments. The goal is to maximize the amount of information gained about the temporal activations of the target sounds. For each unlabeled audio recording, we use a prediction model to derive a probability curve used to guide annotation. The prediction model is initially pre-trained on available annotated sound event data with classes that are disjoint from the classes in the unlabeled dataset. The prediction model then gradually adapts to the annotations provided by the annotator in an active learning loop. We derive query segments to guide the weak label annotator towards strong labels, using change point detection on these probabilities. We show that it is possible to derive strong labels of high quality with a limited annotation budget, and show favorable results for A-CPD when compared to two baseline query segment strategies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_08525 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | From Weak to Strong Sound Event Labels using Adaptive Change-Point Detection and Active Learning Martinsson, John Mogren, Olof Sandsten, Maria Virtanen, Tuomas Sound Machine Learning Audio and Speech Processing We propose an adaptive change point detection method (A-CPD) for machine guided weak label annotation of audio recording segments. The goal is to maximize the amount of information gained about the temporal activations of the target sounds. For each unlabeled audio recording, we use a prediction model to derive a probability curve used to guide annotation. The prediction model is initially pre-trained on available annotated sound event data with classes that are disjoint from the classes in the unlabeled dataset. The prediction model then gradually adapts to the annotations provided by the annotator in an active learning loop. We derive query segments to guide the weak label annotator towards strong labels, using change point detection on these probabilities. We show that it is possible to derive strong labels of high quality with a limited annotation budget, and show favorable results for A-CPD when compared to two baseline query segment strategies. |
| title | From Weak to Strong Sound Event Labels using Adaptive Change-Point Detection and Active Learning |
| topic | Sound Machine Learning Audio and Speech Processing |
| url | https://arxiv.org/abs/2403.08525 |