Table 6 Dataset specifications
From: Bilingual video captioning model for enhanced video retrieval
Dataset split | # of videos | # of captions | # of captions used (English experiment) | # of captions used (Arabic experiment) |
|---|---|---|---|---|
Training | 1200 | 49,142 | 47,472 | 34,477 |
Validation | 100 | 3990 | 3870 | 2836 |
Testing | 670 | 27,695 | 27,695 | 27,695 |
Total | 1970 | 80,827 | 79,037 | 65,008 |