Video and Text Summarization Using VDAN and RNN

Video and Text Summarization Using VDAN and RNN
Author(s):
JOYS PRINCIA A, Miss. Sangeetha Priya, KALAI SELVI J, RITHI AFRA J, RUKSHANA S
Keywords:
VDAN, RNN
Abstract
The main intent of this project is to develop a video and text summarizer. The videos in these days are so quite long. It is difficult for people with hectic work schedule to find time to watch the long videos. Thus a summarizer will help people in getting the gist immediately. The video summarization is done with the help of Visually-Guided Document Attention Network (VDAN).The motive of this network is to extract the textual and visual features. The extraction of visual features is done with the help of Convolutional Neural Network(CNN).The extraction of textual features is done with the help of document level encoding. It also contains Gated Recurrent Unit (GRU).Based on the visual and textual features extracted, the agent decides the corresponding action. The three sets of actions are accelerate, decelerate and do nothing. The text summarization part is done with the help of Recurrent Neural Network. It follows an encoder-decoder architecture. It also makes use of Long Short Term Memory (LSTM) to keep track of the previous observations. Thus at the end the summarized video and text are available.
Article Details
Unique Paper ID: 152248 Publication Volume & Issue: Volume 8, Issue 2 Page(s): 780 - 786
Article Preview & Download

Share This Article

NCSST-2021

AICTE Sponsored National Conference on Smart Systems and Technologies

Last Date: 25th November 2021

LATEST INNOVATION’S AND FUTURE TRENDS IN MANAGEMENT

Last Date: 7th November 2021