Beyond FVD: Enhanced Evaluation Metrics for Video Generation Quality
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_ | 1866917797344313344 |
|---|---|
| author | Luo, Ge Ya Favero, Gian Mario Luo, Zhi Hao Jolicoeur-Martineau, Alexia Pal, Christopher |
| author_facet | Luo, Ge Ya Favero, Gian Mario Luo, Zhi Hao Jolicoeur-Martineau, Alexia Pal, Christopher |
| contents | The Fréchet Video Distance (FVD) is a widely adopted metric for evaluating video generation distribution quality. However, its effectiveness relies on critical assumptions. Our analysis reveals three significant limitations: (1) the non-Gaussianity of the Inflated 3D Convnet (I3D) feature space; (2) the insensitivity of I3D features to temporal distortions; (3) the impractical sample sizes required for reliable estimation. These findings undermine FVD's reliability and show that FVD falls short as a standalone metric for video generation evaluation. After extensive analysis of a wide range of metrics and backbone architectures, we propose JEDi, the JEPA Embedding Distance, based on features derived from a Joint Embedding Predictive Architecture, measured using Maximum Mean Discrepancy with polynomial kernel. Our experiments on multiple open-source datasets show clear evidence that it is a superior alternative to the widely used FVD metric, requiring only 16% of the samples to reach its steady value, while increasing alignment with human evaluation by 34%, on average. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_05203 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Beyond FVD: Enhanced Evaluation Metrics for Video Generation Quality Luo, Ge Ya Favero, Gian Mario Luo, Zhi Hao Jolicoeur-Martineau, Alexia Pal, Christopher Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning The Fréchet Video Distance (FVD) is a widely adopted metric for evaluating video generation distribution quality. However, its effectiveness relies on critical assumptions. Our analysis reveals three significant limitations: (1) the non-Gaussianity of the Inflated 3D Convnet (I3D) feature space; (2) the insensitivity of I3D features to temporal distortions; (3) the impractical sample sizes required for reliable estimation. These findings undermine FVD's reliability and show that FVD falls short as a standalone metric for video generation evaluation. After extensive analysis of a wide range of metrics and backbone architectures, we propose JEDi, the JEPA Embedding Distance, based on features derived from a Joint Embedding Predictive Architecture, measured using Maximum Mean Discrepancy with polynomial kernel. Our experiments on multiple open-source datasets show clear evidence that it is a superior alternative to the widely used FVD metric, requiring only 16% of the samples to reach its steady value, while increasing alignment with human evaluation by 34%, on average. |
| title | Beyond FVD: Enhanced Evaluation Metrics for Video Generation Quality |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2410.05203 |