SORA Explained

YOUR LINK HERE:

http://youtube.com/watch?v=MySoob8j35I

In this session, we're diving into Sora, the amazing new text-to-video model by OpenAI which we'll compare with Google's Lumiere. • About the research: • ------------------------------- • Sora is a scalable, generalist video generation model employing a diffusion transformer architecture, leveraging visual patches inspired by LLMs to generate diverse videos demonstrating emerging simulation capabilities such as 3D consistency, long-term coherence, and interactive behaviors. Lumiere is another text-to-video diffusion model employing a Space-Time U-Net architecture that enables simultaneous generation of temporally coherent, realistic, and diversified videos. • 💡 Project Pages: • Sora: https://openai.com/sora • Lumiere: https://lumiere-video.github.io • 🔬 Sora: https://openai.com/research/video-gen... • Lumiere: A Space-Time Diffusion Model for Video Generation: https://arxiv.org/abs/2401.12945 • 📝 Sora: OpenAI Research • Lumiere: Omer Bar-Tal, Hila Chefer, Omer Tov, Charles Herrmann, Roni Paiss, Shiran Zada, Ariel Ephrat, Junhwa Hur, Guanghui Liu, Amit Raj, Yuanzhen Li, Michael Rubinstein, Tomer Michaeli, Oliver Wang, Deqing Sun, Tali Dekel, Inbar Mosseri • Read also: • ---------------- • 📰 The Deep Dive. Follow the latest AI research and industry trends. https://unifyai.substack.com/ • 📖 Blogs. Dive into the AI deployment stack. https://unify.ai/blog • Follow us: • ---------------- • Website: https://unify.ai • Github: https://github.com/unifyai/ • Discord: / discord • Twitter: / letsunifyai • #ai #openai #lumiere #sora #google #videogeneration

#############################

New on site