DeCoR: Defy Knowledge Forgetting by Predicting Earlier Audio Codes
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866913225398812672 |
|---|---|
| author | Jiang, Xilin Li, Yinghao Aaron Mesgarani, Nima |
| author_facet | Jiang, Xilin Li, Yinghao Aaron Mesgarani, Nima |
| contents | Lifelong audio feature extraction involves learning new sound classes incrementally, which is essential for adapting to new data distributions over time. However, optimizing the model only on new data can lead to catastrophic forgetting of previously learned tasks, which undermines the model's ability to perform well over the long term. This paper introduces a new approach to continual audio representation learning called DeCoR. Unlike other methods that store previous data, features, or models, DeCoR indirectly distills knowledge from an earlier model to the latest by predicting quantization indices from a delayed codebook. We demonstrate that DeCoR improves acoustic scene classification accuracy and integrates well with continual self-supervised representation learning. Our approach introduces minimal storage and computation overhead, making it a lightweight and efficient solution for continual learning. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2305_18441 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | DeCoR: Defy Knowledge Forgetting by Predicting Earlier Audio Codes Jiang, Xilin Li, Yinghao Aaron Mesgarani, Nima Audio and Speech Processing Machine Learning Sound Lifelong audio feature extraction involves learning new sound classes incrementally, which is essential for adapting to new data distributions over time. However, optimizing the model only on new data can lead to catastrophic forgetting of previously learned tasks, which undermines the model's ability to perform well over the long term. This paper introduces a new approach to continual audio representation learning called DeCoR. Unlike other methods that store previous data, features, or models, DeCoR indirectly distills knowledge from an earlier model to the latest by predicting quantization indices from a delayed codebook. We demonstrate that DeCoR improves acoustic scene classification accuracy and integrates well with continual self-supervised representation learning. Our approach introduces minimal storage and computation overhead, making it a lightweight and efficient solution for continual learning. |
| title | DeCoR: Defy Knowledge Forgetting by Predicting Earlier Audio Codes |
| topic | Audio and Speech Processing Machine Learning Sound |
| url | https://arxiv.org/abs/2305.18441 |