Synthetic optical flows generated from single-image depth maps let a simple two-stream network beat prior unsupervised video object segmentation methods on all tested benchmarks.
Dual pro- totype attention for unsupervised video object segmentation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
DepthFlow: Exploiting Depth-Flow Structural Correlations for Unsupervised Video Object Segmentation
Synthetic optical flows generated from single-image depth maps let a simple two-stream network beat prior unsupervised video object segmentation methods on all tested benchmarks.