Wan Streamer is a new end-to-end multimodal foundation model built for real-time, full-duplex audio-visual interaction within a single Transformer architecture. Unlike existing ...
·
Researchers have introduced Sparse Context, a method for dramatically reducing the computational cost of reference-conditioned image generation in diffusion models. The core ins...
Researchers have introduced FlowBender, a training framework for conditional flow and diffusion models that treats a model's own alignment error as a direct input during trainin...
·
Lvmin Zhang, a Stanford University graduate student, created ControlNet, an AI system designed to give artists finer-grained control over AI-generated imagery. Rather than relyi...
Un-0, an image-generation system developed by Databricks' former AI chief, has been unveiled as a demonstration of technology that could reduce artificial intelligence's power c...
·
A CVPR 2025 image super-resolution model called AdcSR has been converted for Apple's Core AI stack, enabling 4× upscaling to run fully on-device — including on iPhone — without ...
·
Researchers from Zhejiang University and ByteDance have introduced Gen3R, a method that bridges foundational reconstruction models and video diffusion models to generate 3D scen...
Researchers from Stability AI and the University of Tübingen have introduced Arbor, a method for controllable 3D asset generation that uses typed geometric constraint meshes as ...
·
Researchers from CityUHK, Stanford, Cornell Tech, and UT Austin have introduced MeshFlow, a diffusion-based generative model that produces 3D triangle meshes directly as unorder...
Researchers from the University of Hong Kong, Alibaba Group, Horizon Robotics, and Ant Group have introduced SCOPE (Scale-Consistent One-Pass Estimation of 3D Geometry), a deep-...
·
