GEB+: A Benchmark for Generic Event Boundary Captioning, Grounding and Retrieval

1 Apr 2022  ยท  Yuxuan Wang, Difei Gao, Licheng Yu, Stan Weixian Lei, Matt Feiszli, Mike Zheng Shou ยท

Cognitive science has shown that humans perceive videos in terms of events separated by the state changes of dominant subjects. State changes trigger new events and are one of the most useful among the large amount of redundant information perceived. However, previous research focuses on the overall understanding of segments without evaluating the fine-grained status changes inside. In this paper, we introduce a new dataset called Kinetic-GEB+. The dataset consists of over 170k boundaries associated with captions describing status changes in the generic events in 12K videos. Upon this new dataset, we propose three tasks supporting the development of a more fine-grained, robust, and human-like understanding of videos through status changes. We evaluate many representative baselines in our dataset, where we also design a new TPD (Temporal-based Pairwise Difference) Modeling method for visual difference and achieve significant performance improvements. Besides, the results show there are still formidable challenges for current methods in the utilization of different granularities, representation of visual difference, and the accurate localization of status changes. Further analysis shows that our dataset can drive developing more powerful methods to understand status changes and thus improve video level comprehension. The dataset is available at https://github.com/showlab/GEB-Plus

PDF Abstract

Datasets


Introduced in the Paper:

Kinetics-GEB+

Used in the Paper:

Kinetics

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Boundary Captioning Kinetics-GEB+ ActBERT-revised CIDEr 74.71 # 1
SPICE 19.52 # 1
ROUGE-L 28.15 # 1
Text to Video Retrieval Kinetics-GEB+ FROZEN-revised mAP 23.39 # 1
Boundary Grounding Kinetics-GEB+ FROZEN-revised F1@0.1s 4.28 # 1
F1@0.2s 8.54 # 1
F1@0.5s 18.33 # 1
F1@1.0s 31.04 # 1
F1@1.5s 40.48 # 1
F1@2.0s 47.86 # 1
F1@2.5s 54.81 # 1
F1@3.0s 61.45 # 1
F1@Avg 33.35 # 1
Text to Video Retrieval Kinetics-GEB+ FROZEN-revised (two-stream) text-to-video R@1 12.8 # 1
text-to-video R@5 34.81 # 1
text-to-video R@10 45.66 # 1
text-to-video R@50 68.1 # 1

Methods


No methods listed for this paper. Add relevant methods here