Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2408.02209 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909280392708096 |
|---|---|
| author | Khramtsova, Ekaterina Baktashmotlagh, Mahsa Zuccon, Guido Wang, Xi Salzmann, Mathieu |
| author_facet | Khramtsova, Ekaterina Baktashmotlagh, Mahsa Zuccon, Guido Wang, Xi Salzmann, Mathieu |
| contents | Accurately estimating model performance poses a significant challenge, particularly in scenarios where the source and target domains follow different data distributions. Most existing performance prediction methods heavily rely on the source data in their estimation process, limiting their applicability in a more realistic setting where only the trained model is accessible. The few methods that do not require source data exhibit considerably inferior performance. In this work, we propose a source-free approach centred on uncertainty-based estimation, using a generative model for calibration in the absence of source data. We establish connections between our approach for unsupervised calibration and temperature scaling. We then employ a gradient-based strategy to evaluate the correctness of the calibrated predictions. Our experiments on benchmark object recognition datasets reveal that existing source-based methods fall short with limited source sample availability. Furthermore, our approach significantly outperforms the current state-of-the-art source-free and source-based methods, affirming its effectiveness in domain-invariant performance estimation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_02209 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Source-Free Domain-Invariant Performance Prediction Khramtsova, Ekaterina Baktashmotlagh, Mahsa Zuccon, Guido Wang, Xi Salzmann, Mathieu Computer Vision and Pattern Recognition Accurately estimating model performance poses a significant challenge, particularly in scenarios where the source and target domains follow different data distributions. Most existing performance prediction methods heavily rely on the source data in their estimation process, limiting their applicability in a more realistic setting where only the trained model is accessible. The few methods that do not require source data exhibit considerably inferior performance. In this work, we propose a source-free approach centred on uncertainty-based estimation, using a generative model for calibration in the absence of source data. We establish connections between our approach for unsupervised calibration and temperature scaling. We then employ a gradient-based strategy to evaluate the correctness of the calibrated predictions. Our experiments on benchmark object recognition datasets reveal that existing source-based methods fall short with limited source sample availability. Furthermore, our approach significantly outperforms the current state-of-the-art source-free and source-based methods, affirming its effectiveness in domain-invariant performance estimation. |
| title | Source-Free Domain-Invariant Performance Prediction |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2408.02209 |