A presenter is discussing advancements in multimodal-guided video generation at a conference, using a slide to illustrate their points. The slide showcases various projects, including "MovieFactory," "CoDi," "MM-Diffusion," and "NEXT-GPT," each with annotated diagrams and brief descriptions. The audience attentively follows along, highlighting the engaged and collaborative atmosphere typical of specialized tech conferences. The slide is titled "Multimodal-Guided Video Generation: More Works," underlining the focus on recent developments and ongoing research in this cutting-edge field.
Text transcribed from the image:
Multimodal-Guided Video Generation: More Works
Sep 1: Spatial Fl
Step 2: Tempiral Training
Composable
Cendening
Stam 2
www
Cecoration
(b)
Cendoning Alg
Coding
Bingle Draining
L
Presined and Final Moddle
Added and Table Module
MovieFactory (Zhu et al.)
"MovieFactory: Automatic Movie Creation from Text using Large Generative Models
for Language and Images," arXiv 2023.
(0)
CODI (Tang et al.)
ଅ
"Any-to-Any Generation via Composable Diffusion," NeuriPS 2023.
M
00
Audo Branch and Sp
Ereding
MM-Diffusion (Ruan et al.)
"MM-Diffusion: Learning Multi-Modal Diffusion Models for Joint Audio and Video
Generation," CVPR 2023.
Xing et al., "A Survey on Video Diffusion Models," arXiv 2023.
Video
More modales
掴
Above]
the
Projesi
LIM
Audio Output
Encoding
A
Alignment
Generation
NEXT-GPT (Wu et al.)
"NEXT-GPT: Any-to-Any Multimodal LLM," arXiv 2023.
Copyright Mike Shou, NUS
160
Я ТЕХОЛ Е