Inverse-designed metasurface combines full-Stokes polarimetry and wavefront sensing in a single continuous aperture, achieving 0.046 mean polarization reconstruction error with neural network assistance.
Autoencoder with recurrent neural networks for video forgery detection
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
Video forgeries are detectable via binary classification on multimedia stream descriptors without pixel analysis.
Implementation of DQN and modified DQN on a custom 2D track reports the modified version achieving around 40 average reward after 1000 episodes, 60% higher than standard DQN.
citing papers explorer
-
Inverse designed full-Stokes polarimetric metasurface with simultaneous wavefront sensing for visible light
Inverse-designed metasurface combines full-Stokes polarimetry and wavefront sensing in a single continuous aperture, achieving 0.046 mean polarization reconstruction error with neural network assistance.
-
We Need No Pixels: Video Manipulation Detection Using Stream Descriptors
Video forgeries are detectable via binary classification on multimedia stream descriptors without pixel analysis.
-
Enhanced Deep Q-Learning for 2D Self-Driving Cars: Implementation and Evaluation on a Custom Track Environment
Implementation of DQN and modified DQN on a custom 2D track reports the modified version achieving around 40 average reward after 1000 episodes, 60% higher than standard DQN.