EyeSeg adds an uncertainty head to an eye segmentation network, proving that the predicted variance equals the feature distance to the class center under known labels, and uses this score to filter unreliable predictions.
Encoder-decoder with atrous separable convolution for se- mantic image segmentation
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
-
EyeSeg: An Uncertainty-Aware Eye Segmentation Framework for AR/VR
EyeSeg adds an uncertainty head to an eye segmentation network, proving that the predicted variance equals the feature distance to the class center under known labels, and uses this score to filter unreliable predictions.