CycDA: Unsupervised Cycle Domain Adaptation to Learn from Image to Video.

Wei Lin Anna Kukleva Kunyang Sun Horst Possegger Hilde Kuehne Horst Bischof

Published in: ECCV (3) (2022)

Keyphrases

domain adaptation
unsupervised manner
input image
image features
labeled data
feature representation
video sequences
semi supervised
cross domain
machine learning
similarity measure
sentiment classification
transfer knowledge