List your company

Video-grounded Dialogue Generation on VSTAR (test)

0.092BLEU-1AVDT ViT
0.07120.07660.0820.0874May 30, 2023

Evaluation results

MethodLinks
AVDT ViTVisual Backbone=ViT2023.050.0920.0410.0230.0140.0410.0820.144
AVDT w/o seg (3 frames)segmentation=none, number of frames=32023.050.090.040.0220.0140.0410.0810.139
AVDT2023.050.0890.040.0220.0140.0410.0820.145
AVDT Faster-RCNNVisual Backbone=Faster-RCNN2023.050.0890.040.0220.0140.040.080.137
AVDT w/o seg (1 frame)segmentation=none, number of frames=12023.050.0870.0390.0210.0130.0390.0770.126
AVDT w/o seg text-onlysegmentation=none, text-only=true2023.050.0850.0370.0210.0130.0370.0750.126
AVDT w/o seg query-onlysegmentation=none, query-only=true2023.050.0820.0350.0190.0110.0370.0730.119
OpenViDial coarse2023.050.0750.0260.0130.0060.0350.0630.066
RLM2023.050.0720.0320.0170.010.0320.0610.079