Visual Storytelling On Vist
評価指標
BLEU-4
CIDEr
METEOR
ROUGE-L
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
比較表
| モデル名 | BLEU-4 | CIDEr | METEOR | ROUGE-L |
|---|---|---|---|---|
| knowledgeable-storyteller-a-commonsense | 12.8 | 12.1 | 35.2 | 29.9 |
| plot-and-rework-modeling-storylines-for | 7.65 | - | 31.6 | - |
| no-metrics-are-perfect-adversarial-reward | 14.1 | 9.4 | 35 | 29.5 |
| visual-storytelling-with-hierarchical-bert | 15.4 | - | 36.5 | - |
| keep-it-consistent-topic-aware-storytelling | 14.6 | 9.2 | 35.7 | 31 |
| sentistory-a-multi-layered-sentiment-aware | 14.8 | 10.1 | 35.7 | 30.2 |
| what-makes-a-good-story-designing-composite | 14.4 | 6.7 | 35.2 | 30.1 |
| what-makes-a-good-story-designing-composite | 9.8 | 5.9 | 30.1 | 25.1 |
| what-makes-a-good-story-designing-composite | 12.4 | 8.6 | 33.9 | 29.9 |
| two-heads-are-better-than-one-hypergraph | 16.7 | 14.1 | 37.8 | - |
| transitional-adaptation-of-pretrained-models | - | 8.3 | 34.1 | 30.2 |
| glac-net-glocal-attention-cascading-networks | - | - | 30.14 | - |
| no-metrics-are-perfect-adversarial-reward | 13.7 | 8.7 | 34.8 | 29.7 |
| no-metrics-are-perfect-adversarial-reward | 14 | 9 | 35 | 29.5 |
| hide-and-tell-learning-to-bridge-photo | 14.7 | 10 | 35.6 | 29.7 |
| bert-hlstms-bert-and-hierarchical-lstms-for | - | 8.37 | - | - |
| commonsense-knowledge-aware-concept-selection | 13 | 11 | 36.1 | 30.7 |
| diverse-and-relevant-visual-storytelling-with | 14.8 | 8.6 | 35.6 | 30.2 |
| hierarchical-memory-decoder-for-visual | 14.1 | - | 35.5 | - |
| storytelling-from-an-image-stream-using-scene | 14.7 | 9.8 | 35.8 | 29.9 |
| what-makes-a-good-story-designing-composite | 13.6 | 9.1 | 35.2 | 29.3 |
| hierarchically-structured-reinforcement | 12.32 | 10.71 | 35.23 | 30.84 |
| vision-transformer-based-model-for-describing | 12.3 | 4.4 | 35.4 | 31 |
| transitional-adaptation-of-pretrained-models | - | 13.8 | 37.2 | 33.1 |
| imagine-reason-and-write-visual-storytelling | 15.4 | 11.0 | 35.6 | 29.6 |
| visual-storytelling-via-predicting-anchor | 14 | 9.9 | 35.5 | 30 |
| aog-lstm-an-adaptive-attention-neural-network | 12.9 | 12.0 | 36.0 | 30.1 |
| hierarchically-attentive-rnn-for-album | - | 7.38 | 33.94 | 29.82 |
| what-makes-a-good-story-designing-composite | 14.3 | 7.2 | 34.8 | 30 |
| coherent-visual-storytelling-via-parallel-top | 15.2 | 11.5 | 36.5 | 30.8 |
| informative-visual-storytelling-with-cross | 14.3 | 9 | 35.5 | 30.2 |
| contextualize-show-and-tell-a-neural-visual | 12.7 | 5.1 | 34.4 | 29.2 |
| bert-hlstms-bert-and-hierarchical-lstms-for | - | 7.98 | - | - |