Today in Generative Media

Media fights, negotiates; Artists ramp up AI dissent; Meta's multimodal Chameleon

• The media bosses fighting back against AI — and the ones cutting deals (Washington Post)

• As AI companies lean on creative industries for training data, artists ramp up their dissent (Fortune)

• Meta introduces Chameleon, a state-of-the-art multimodal model (VentureBeat)

• Adobe’s controversial marketing is fueling a war with creatives over AI (Fast Company)

• Announcing our partnership with ElevenLabs for text-to-speech and voice API (Synthesia)

• Vidu4D: Single Generated Video to High-Fidelity 4D Reconstruction with Dynamic Gaussian Surfels. (project page) Paper on arXiv.

• Hybrid Mesh-Gaussian Head Avatar for High-Fidelity Rendering and Head Editing (project page)

• Video Diffusion Models are Training-free Motion Interpreter and Controller (project page)

• NieR: Normal-Based Lighting Scene Rendering (project page)

• GSDeformer: Direct Cage-based Deformation for 3D Gaussian Splatting (project page)

• Look Once to Hear: Target Speech Hearing with Noisy Examples (CHI 2024, best paper honorable mention, arXiv)

• HDR-GS: Efficient High Dynamic Range Novel View Synthesis at 1000x Speed via Gaussian Splatting (arXiv)

• Gaussian Time Machine: A Real-Time Rendering Methodology for Time-Variant Appearances (arXiv)

• iVideoGPT: Interactive VideoGPTs are Scalable World Models (arXiv)

• D-MiSo: Editing Dynamic 3D Scenes using Multi-Gaussians Soup (arXiv)

• A Dynamic By-example BTF Synthesis Scheme (arXiv)

• PIGEON: Predicting Image Geolocations (arXiv)

• PyGS: Large-scale Scene Representation with Pyramidal 3D Gaussian Splatting (arXiv)