Video Generation
Video generation is an important task in the field of computer vision, aiming to automatically generate high-quality video content through algorithms. The goal is to use deep learning models, such as Generative Adversarial Networks (GANs) and Variational Autoencoders (VAEs), to create coherent and realistic dynamic videos from static images, text descriptions, or other video clips. Video generation technology has broad application value in areas like film production, virtual reality, game development, and advertising creativity, significantly enhancing the efficiency and innovation of content creation.
BAIR Robot Pushing
MAGVIT
How2Sign
Kinetics-600 12 frames, 128x128
DVD-GAN
Kinetics-600 12 frames, 64x64
MAGVIT
Kinetics-600 48 frames, 64x64
DVD-GAN
LAION-400M
MSR-VTT
VideoAssembler (Zero-Shot, 256x256, class-conditional)
Sky Time-lapse
Taichi
DIGAN (256x256)
TrailerFaces
PG-SWGAN-3D
UCF-101
FIFO-Diffusion
UCF-101 16 frames, 128x128, Unconditional
TGANv2 (2020)
UCF-101 16 frames, 64x64, Unconditional
Video Diffusion Model
UCF-101 16 frames, Unconditional, Single GPU
TGAN-F
UCF101
YouTube Driving
StyleSV