New object-detection and saliency descriptors, combined with uncertainty-weighted feature hallucination, improve RGB-only action recognition on multiple video benchmarks.
In: 2008 Fifth International Conference on Computer Graphics, Imaging and Visualisation, pp
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Feature Hallucination for Self-supervised Action Recognition
New object-detection and saliency descriptors, combined with uncertainty-weighted feature hallucination, improve RGB-only action recognition on multiple video benchmarks.