Ada-SwinBERT: Adaptive Token Selection for Efficient Video Captioning with Online Self-Distillation.

Qianwen Cao Heyan Huang Minpeng Liao Xianling Mao

Published in: ICME (2023)

Keyphrases

real time
online video
video streams
video sequences
computationally expensive
video surveillance
content aware
real time video
digital video
video clips
video content
video data
online learning
multimedia
cost effective
space time
computationally efficient
lightweight
scalable video
neural network