Test-time training on unlabeled videos from a new event improves fake news video detection accuracy on FakeSV, with gains of 2.48% (event split) and 3.32% (temporal split) over prior state of the art.
In: 2017 ieee international conference on acoustics, speech and signal processing (icassp)
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
-
T$^\text{3}$SVFND: Towards an Evolving Fake News Detector for Emergencies with Test-time Training on Short Video Platforms
Test-time training on unlabeled videos from a new event improves fake news video detection accuracy on FakeSV, with gains of 2.48% (event split) and 3.32% (temporal split) over prior state of the art.