BAM-DETR: Boundary-Aligned Moment Detection Transformer for Temporal Sentence Grounding in Videos

Benchmark Model Rank Results
moment-retrieval-on-charades-staBAM-DETR#12R@1 IoU=0.5: 59.95R@1 IoU=0.7: 39.38
moment-retrieval-on-qvhighlightsBAM-DETR (w/ audio)#9mAP: 46.91R@1 IoU=0.5: 64.07R@1 IoU=0.7: 48.12mAP@0.5: 65.61
moment-retrieval-on-qvhighlightsBAM-DETR (w/ PT ASR Captions)#10mAP: 46.67R@1 IoU=0.5: 63.88R@1 IoU=0.7: 47.92mAP@0.5: 66.33
moment-retrieval-on-qvhighlightsBAM-DETR#13mAP: 45.36R@1 IoU=0.5: 62.71R@1 IoU=0.7: 48.64mAP@0.5: 64.57
natural-language-moment-retrieval-on-tacosBAM-DETR#5R@1,IoU=0.3: 56.69R@1,IoU=0.5: 41.54R@1,IoU=0.7: 26.77mIoU: 39.31