NDCG (x 100): 57.59MRR (x 100): 63.2R@1: 49.63R@5: 79.75
Papers with Code · CC BY-SA 4.0
Visual Dialog
Visual Dialog requires an AI agent to hold a meaningful dialog with humans in natural, conversational language about visual content. Specifically, given an image, a dialog history, and a follow-up question about the image, the task is to answer the question.
논문
NDCG (x 100): 56.32MRR (x 100): 63.23R@1: 49.25R@5: 80.23
NDCG (x 100): 57.20MRR (x 100): 69.3R@1: 55.65R@5: 86.73
NDCG (x 100): 57.32MRR (x 100): 62.20R@1: 47.90R@5: 80.43
NDCG (x 100): 56.64MRR (x 100): 63.49R@1: 49.85R@5: 80.63
NDCG (x 100): 58.1MRR (x 100): 58.8R@1: 44.15R@5: 76.88
NDCG (x 100): 57.17MRR (x 100): 64.22R@1: 50.88R@5: 80.63
NDCG (x 100): 59.37MRR (x 100): 64.84R@1: 51.45R@5: 81.12
NDCG (x 100): 52.82MRR (x 100): 61.37R@1: 47.33R@5: 77.98
NDCG (x 100): 55.59MRR (x 100): 63.03R@1: 49.03R@5: 80.40
NDCG (x 100): 54.70MRR (x 100): 61.50R@1: 47.55R@5: 78.10