arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.03315cs.CVcs.LGeess.IV
收藏

VATEX字幕挑战赛2020的多模态特征融合与特征关注

Multi-modal Feature Fusion with Feature Attention for VATEX Captioning Challenge 2020

Samsung Research China - Beijing (SRC-B)(三星中国研究院-北京(SRC-B));Peking University(北京大学)

Ke Lin, Zhuoxin Gan, Liwei Wang

摘要:

本报告描述了我们的VATEX字幕挑战2020模型。首先,从多个领域收集信息,我们提取运动,外观,语义和音频特征。然后设计了一个特征关注模块,在解码时关注不同的特征。我们采用两种类型的解码器,自顶向下和X-LAN和集成这些模型得到最终的结果。所提出的方法优于官方基线有显着的差距。我们在英语和中文私人测试集上分别获得了76.0和50.0的CIDEr。我们在英语和中文私人测试排行榜上排名第二。

英文摘要:

This report describes our model for VATEX Captioning Challenge 2020. First, to gather information from multiple domains, we extract motion, appearance, semantic and audio features. Then we design a feature attention module to attend on different feature when decoding. We apply two types of decoders, top-down and X-LAN and ensemble these models to get the final result. The proposed method outperforms official baseline with a significant gap. We achieve 76.0 CIDEr and 50.0 CIDEr on English and Chinese private test set. We rank 2nd on both English and Chinese private test leaderboard.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑