PulseAugur / Brief
EN
LIVE 01:01:24

Brief

last 24h
[2/2] 222 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. RESEARCH · Mastodon — fosstodon.org 한국어(KO) · · [2 sources]

    NVIDIA's SANA-WM world model has been released as open source by Victor M (@victormustar). It is a camera-conditional world model that runs on a single GPU, and it is introduced as processing 60 seconds of 720p video in 34 seconds with 2.6B parameters. World models, simulation, agents

    NVIDIA has open-sourced its SANA-WM world model, which can generate a minute of 720p video from a single image. This camera-conditional model operates on a single GPU and processes 60 seconds of video in 34 seconds. The release is considered a practical advancement for research in world models, simulations, and agents. AI

    NVIDIA's SANA-WM world model has been released as open source by Victor M (@victormustar). It is a camera-conditional world model that runs on a single GPU, and it is introduced as processing 60 seconds of 720p video in 34 seconds with 2.6B parameters. World models, simulation, agents

    IMPACT Enables researchers and developers to experiment with and build upon a new world model for video generation.

  2. TOOL · Hugging Face Trending Models English(EN) ·

    Efficient-Large-Model/SANA-WM_bidirectional

    Researchers have released SANA-WM, an open-source world model capable of generating minute-long videos at 720p resolution. This diffusion transformer model utilizes a hybrid linear attention mechanism and a dual-branch architecture for precise camera control. The model also incorporates a two-stage generation pipeline with a refiner for enhanced quality and temporal consistency, and it was trained using a robust annotation pipeline with metric-scale 6-DoF camera poses. AI

    IMPACT Enables creation of longer, high-fidelity videos with precise camera control, potentially impacting content generation and simulation.