The DKU Audio-Visual Wake Word Spotting System for the 2021 MISP Challenge.

Ming Cheng Haoxu Wang Yechen Wang Ming Li

Published in: ICASSP (2022)

Keyphrases

audio visual
word spotting
multi modal
visual information
text processing
multi stream
dynamic time warping
document images
visual data
handwriting recognition
multimedia
handwritten documents
optical character recognition
image processing
image content
data sets
image sequences
metadata