Closed Mikeltec closed 2 weeks ago
Code included in the PreProcessing-FrameExtraction notebook to help the 3D CNN model learn to identify key elements and actions in GIFs even if they are flipped or slightly rotated, leading to more accurate and generalizable descriptions.
Code included in the PreProcessing-FrameExtraction notebook to help the 3D CNN model learn to identify key elements and actions in GIFs even if they are flipped or slightly rotated, leading to more accurate and generalizable descriptions.