DID YOU KNOW? DTIC has over 3.5 million final reports on DoD funded research, development, test, and evaluation activities available to our registered users. Click
HERE to register or log in.
Accession Number:
AD1049685
Title:
Improving LSTM-based Video Description with Linguistic Knowledge Mined from Text
Report Date:
2016-11-29
Abstract:
This paper investigates how linguistic knowledge mined from large text corpora can aid the generation of natural language descriptions of videos. Specifically, we integrate both a neural language model and distributional semantic strained on large text corpora into a recent LSTM-based architecture for video description. We evaluate our approach on a collection of Youtube videos as well as two large movie description datasets showing significant improvements in grammaticality while modestly improving descriptive quality.
Document Type:
Conference:
Journal:
Pages:
6
File Size:
1.73MB
N/A
(N/A);
IIS-1427425, IIS-1212798
( IIS1427425, IIS1212798);
N/A
(N/A);
N00014-11-1-010
( N00014111010);
FA8750-13-2-0026
( FA87501320026);
Contracts:
Grants:
Distribution Statement:
Approved For Public Release
