Multi-Sentence Video Captioning using Content-oriented Beam Searching and Multi-stage Refining Algorithm
作者：
Highlights：
• A new multi-sentence video captioning algorithm is proposed.
• A new content-oriented beam search approach and a multi-stage refining method are used.
• A new neural network is proposed to measure the relevance between a sentence and a video.
• The structural dictionary of sentences is used to update the probabilities of words.
• An object detector is used to enhance the effectiveness of the algorithm.
摘要
•A new multi-sentence video captioning algorithm is proposed.•A new content-oriented beam search approach and a multi-stage refining method are used.•A new neural network is proposed to measure the relevance between a sentence and a video.•The structural dictionary of sentences is used to update the probabilities of words.•An object detector is used to enhance the effectiveness of the algorithm.
论文关键词：Multi-sentence video captioning,Beam search algorithm,Multimodal relevance measure network
论文评审过程：Received 19 December 2019, Revised 18 April 2020, Accepted 12 May 2020, Available online 16 June 2020, Version of Record 16 June 2020.
论文官网地址：https://doi.org/10.1016/j.ipm.2020.102302
