How Knowledge Distillation Mitigates the Synthetic Gap in Fair Face Recognition
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_ | 1866917763872718848 |
|---|---|
| author | Neto, Pedro C. Colakovic, Ivona Karakatič, Sašo Sequeira, Ana F. |
| author_facet | Neto, Pedro C. Colakovic, Ivona Karakatič, Sašo Sequeira, Ana F. |
| contents | Leveraging the capabilities of Knowledge Distillation (KD) strategies, we devise a strategy to fight the recent retraction of face recognition datasets. Given a pretrained Teacher model trained on a real dataset, we show that carefully utilising synthetic datasets, or a mix between real and synthetic datasets to distil knowledge from this teacher to smaller students can yield surprising results. In this sense, we trained 33 different models with and without KD, on different datasets, with different architectures and losses. And our findings are consistent, using KD leads to performance gains across all ethnicities and decreased bias. In addition, it helps to mitigate the performance gap between real and synthetic datasets. This approach addresses the limitations of synthetic data training, improving both the accuracy and fairness of face recognition models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2408_17399 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | How Knowledge Distillation Mitigates the Synthetic Gap in Fair Face Recognition Neto, Pedro C. Colakovic, Ivona Karakatič, Sašo Sequeira, Ana F. Computer Vision and Pattern Recognition Leveraging the capabilities of Knowledge Distillation (KD) strategies, we devise a strategy to fight the recent retraction of face recognition datasets. Given a pretrained Teacher model trained on a real dataset, we show that carefully utilising synthetic datasets, or a mix between real and synthetic datasets to distil knowledge from this teacher to smaller students can yield surprising results. In this sense, we trained 33 different models with and without KD, on different datasets, with different architectures and losses. And our findings are consistent, using KD leads to performance gains across all ethnicities and decreased bias. In addition, it helps to mitigate the performance gap between real and synthetic datasets. This approach addresses the limitations of synthetic data training, improving both the accuracy and fairness of face recognition models. |
| title | How Knowledge Distillation Mitigates the Synthetic Gap in Fair Face Recognition |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2408.17399 |