IG-FIQA: Improving Face Image Quality Assessment through Intra-class Variance Guidance robust to Inaccurate Pseudo-Labels
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909601786494976 |
|---|---|
| author | Kim, Minsoo Nam, Gi Pyo Kim, Haksub Park, Haesol Kim, Ig-Jae |
| author_facet | Kim, Minsoo Nam, Gi Pyo Kim, Haksub Park, Haesol Kim, Ig-Jae |
| contents | In the realm of face image quality assesment (FIQA), method based on sample relative classification have shown impressive performance. However, the quality scores used as pseudo-labels assigned from images of classes with low intra-class variance could be unrelated to the actual quality in this method. To address this issue, we present IG-FIQA, a novel approach to guide FIQA training, introducing a weight parameter to alleviate the adverse impact of these classes. This method involves estimating sample intra-class variance at each iteration during training, ensuring minimal computational overhead and straightforward implementation. Furthermore, this paper proposes an on-the-fly data augmentation methodology for improved generalization performance in FIQA. On various benchmark datasets, our proposed method, IG-FIQA, achieved novel state-of-the-art (SOTA) performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_08256 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | IG-FIQA: Improving Face Image Quality Assessment through Intra-class Variance Guidance robust to Inaccurate Pseudo-Labels Kim, Minsoo Nam, Gi Pyo Kim, Haksub Park, Haesol Kim, Ig-Jae Computer Vision and Pattern Recognition In the realm of face image quality assesment (FIQA), method based on sample relative classification have shown impressive performance. However, the quality scores used as pseudo-labels assigned from images of classes with low intra-class variance could be unrelated to the actual quality in this method. To address this issue, we present IG-FIQA, a novel approach to guide FIQA training, introducing a weight parameter to alleviate the adverse impact of these classes. This method involves estimating sample intra-class variance at each iteration during training, ensuring minimal computational overhead and straightforward implementation. Furthermore, this paper proposes an on-the-fly data augmentation methodology for improved generalization performance in FIQA. On various benchmark datasets, our proposed method, IG-FIQA, achieved novel state-of-the-art (SOTA) performance. |
| title | IG-FIQA: Improving Face Image Quality Assessment through Intra-class Variance Guidance robust to Inaccurate Pseudo-Labels |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.08256 |