[Generative] Align your Latents:High-Resolution Video Synthesis with Latent Diffusion Models (CVPR’23)

Paper : https://openaccess.thecvf.com/content/CVPR2023/html/Blattmann_Align_Your_Latents_High-Resolution_Video_Synthesis_With_Latent_Diffusion_Models_CVPR_2023_paper.htmlAuthorsNvidia + Andreas Blattmann et al, Toronto Univ., CVPR’23 DecMain IdeaLDM을 기반으로 computation & memory 측면에서 효율적인 접근으로 Vido Generation을 했습니다.기존 pre-trained된 image LDM을 활용temporal layer를 image에 align함으로써High Resolution & Long-..

2024.03.08