Login / Signup
Hypothesis Stitcher for End-to-End Speaker-attributed ASR on Long-form Multi-talker Recordings.
Xuankai Chang
Naoyuki Kanda
Yashesh Gaur
Xiaofei Wang
Zhong Meng
Takuya Yoshioka
Published in:
CoRR (2021)
Keyphrases
</>
end to end
automatic speech recognition
audio visual
high bandwidth
speech recognition
admission control
multipath
wireless ad hoc networks
spontaneous speech
congestion control
multimedia
ad hoc networks
content delivery
packet loss rate
acoustic features
speech signal
image sequences