Implementation of SAT (Show, Attend and Tell)
python main.py --epochs [epoch]
python main.py --prev_model [model_name] --inference --img_path ./data/Flicker8k/images/[img_name]
BLEU | BLEU-1 | BLEU-2 | BLEU-3 | BLEU-4 |
---|---|---|---|---|
23.41 | 65.42 | 39.36 | 21.31 | 12.19 |
- Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
(https://arxiv.org/abs/1502.03044) - Flickr8k Dataset
(https://www.kaggle.com/adityajn105/flickr8k https://github.com/AaronCCWong/Show-Attend-and-Tell