Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore Google's latest Text to Video model in this 11-minute technical video that delves into solving temporal inconsistency challenges in video generation. Learn about the innovative space-time UNet architecture and MultiDiffusion design through a comprehensive breakdown of the model's components. Understand temporal consistency issues, examine the text-to-video generation pipeline, and discover how the Space-Time UNet (STUNet) architecture revolutionizes video creation. Follow along as the presenter, an experienced Machine Learning Researcher, analyzes both qualitative and quantitative results, demonstrating the model's effectiveness in generating coherent and high-quality video content from text descriptions.
Syllabus
- Intro
- Temporal Consistency
- Text to Video generation Pipeline
- UNet Overview
- Space-Time UNet or STUNet
- MultiDiffusion
- Qualitative Results
- Quantitative Results
- Conclusion
Taught by
AI Bites