Today in Generative Media

The battle for AI's future; Instagram's custom AI chatbots, Canva acquires Leonardo.ai

News

• Mark Zuckerberg Just Intensified the Battle for AI’s Future (Time)

• Silicon Valley shaken as open-source AI models Llama 3.1 and Mistral Large 2 match industry leaders (VentureBeat)

• Instagram Will Let You Make Custom AI Chatbots—Even Ones Based on Yourself (Wired)

• Canva acquires Leonardo.ai to boost its generative AI efforts (TechCrunch)

• When AI is trained with AI-generated data, it starts spouting gibberish (Futurism)

Research

• Zero Grads: Learning Local Surrogate Losses for Non-Differentiable Graphics (SIGGRAPH 2024, project page)

• PartGLEE: A Foundation Model for Recognizing and Parsing Any Objects (ECCV 2024, project page)

• Text2Place : Affordance Aware Human Guided Placement (ECCV 2024, project page)

• Viewpoint Textual Inversion: Discovering Scene Representations and 3D View Control in 2D Diffusion Models (ECCV 2024, project page)

• Global Structure-from-Motion Revisited (ECCV 2024, project page)

• Stable Audio Open generates variable-length (up to 47s) stereo audio at 44.1kHz from text prompts. (project page)

• Outfit Anyone: Ultra-high quality virtual try-on for Any Clothing and Any Person (project page)

• Stable-Hair: Real-World Hair Transfer via Diffusion Model (project page)

• Cinemo: Consistent and Controllable Image Animation with Motion Diffusion Models (project page)

• HumanVid: Demystifying Training Data for Camera-controllable Human Image Animation (project page)

• Artist: Aesthetically Controllable Text-Driven Stylization without Training (project page)

• Floating No More: Object-Ground Reconstruction from a Single Image (project page)

• Media2Face: Co-speech Facial Animation Generation (project page)