Detecting meaningful events in an untrimmed video is essential for dense video captioning. In this work, we propose a novel and simple model for event sequence generation and explore temporal relationships of the event sequence in the video. The proposed model omits inefficient two-stage proposal generation and directly generates event boundaries conditioned on bi-directional temporal dependency in one pass. Experimental results show that the proposed event sequence generation model can generate more accurate and diverse events within a small number of proposals. For the event captioning, we follow our previous work to employ the intra-event captioning models into our pipeline system. The overall system achieves state-of-the-art performance on the dense-captioning events in video task with 9.894 METEOR score on the challenge testing set.
@article{arxiv.2006.07896,
title = {Team RUC_AIM3 Technical Report at Activitynet 2020 Task 2: Exploring Sequential Events Detection for Dense Video Captioning},
author = {Yuqing Song and Shizhe Chen and Yida Zhao and Qin Jin},
journal= {arXiv preprint arXiv:2006.07896},
year = {2020}
}
Comments
Winner solution in CVPR 2020 Activitynet Dense Video Captioning challenge