GANs for Video Generation

3D Convolutional GAN

Architecture using 3D convolutions to directly process the spatio-temporal volume (height × width × time), naturally capturing temporal relationships in videos.

← Wstecz