🚨Can existing video generative AI 🎥 replace graphic designers in Hollywood 🎬? The answer is NO 🚫 📢 We present VideoPhy, and find that the generated videos lack physical commonsense – making them poor physical world simulators 🤖 Paper: github.com/Hritikbansal/v… 🧵👇

Want to generate toy figures 🏂🐕🦘🦸🪿⛄ standing on your desk using text-to-3D models? We introduce Atlas3D — a plug-and-play algorithm that endows text-to-3D models with standability. Paper: arxiv.org/pdf/2405.18515 Code: github.com/yunuoch/Atlas3D