3D modeling

FilmAgent is an innovative film production technology that uses multi-agent collaboration to achieve automated film production in a virtual 3D space by simulating key roles such as directors, screenwriters, actors and photographers. The main advantage of this technology is that it can reduce manual intervention, improve production efficiency, and reduce error rates. The application of FilmAgent in the field of film production provides creators with an efficient and low-cost solution, especially suitable for small production teams with limited resources. Although there is currently no clear price information, its open source nature makes it widely applicable and popular.

StereoCrafter is an innovative framework that utilizes base models as priors to convert 2D videos into immersive stereoscopic 3D videos through depth estimation and stereoscopic video inpainting techniques. This technology breaks through the limitations of traditional methods and improves the high-fidelity generation performance required for display devices. Key benefits of StereoCrafter include the ability to handle video inputs of different lengths and resolutions, as well as optimized video processing through autoregressive strategies and chunking. In addition, StereoCrafter has developed sophisticated data processing pipelines to reconstruct large-scale, high-quality data sets to support the training process. This framework provides practical solutions for creating immersive content with 3D devices such as Apple Vision Pro and 3D displays, potentially changing the way we experience digital media.

Long Volumetric Video is a new technique for reconstructing long voxel videos from multi-view RGB videos. This technology uses Temporal Gaussian Hierarchy, a novel 4D representation method, to compactly simulate long voxel videos, solving the problems of large memory usage and slow rendering speed of traditional dynamic view synthesis methods when processing long videos. The main advantages of this technology include low training cost, fast rendering speed and low storage usage, and it is the first technology to efficiently process minute-level voxel video data while maintaining high-quality rendering.

Wonder Animation is an innovative product from Autodesk that uses cutting-edge video to 3D scene technology to convert any video sequence into a 3D animated scene to accelerate the animated film production process. This technology is able to process film and editorial sequences containing multiple cuts and different shots (wide angle, medium shot, close-up), and then use AI to reconstruct the scene in 3D space and match each camera's position and motion relationship with the characters and environment. The main advantage of Wonder Animation is its ability to provide artists with complete creative control, giving artists the ability to shape their own vision, unlike other generative AI tools on the market that rely on automated output.

This study proposes a new hierarchical 3D Gaussian representation method for real-time rendering of very large datasets. This method provides excellent visual quality, fast training and real-time rendering capabilities through 3D Gaussian splatting technology. Through layered structures and effective Level-of-Detail (LOD) solutions, distant content can be rendered efficiently and smooth transitions between different levels can be achieved. The technique is able to adapt to available resources, training large scenes through a divide-and-conquer approach, and integrating them into a hierarchy that can be further optimized to improve visual quality when Gaussians are merged into intermediate nodes.

DreamWorld AI is an artificial intelligence and computer vision research and development company focused on building the next generation of AI-driven digital humans. The company's proprietary AI models and algorithms allow users to create, animate and perform real-time full-body digital characters in a variety of styles using only a single-lens device, without the need for suits, markers or special equipment. The platform provides creators with a full-stack AI-driven virtual production workstation, allowing creators to easily produce high-quality virtual character content.

Cascadeur is an independent 3D keyframe animation software that can be used for animation of humanoid characters or other characters. With its AI-assisted tools, you can quickly create key poses, instantly see the physics and adjust secondary movements while maintaining full control.