PixelWizard: Towards Efficient High-Fidelity Video Generation at Ultra-Large Spatial Resolution
This work provides an efficient framework for generating high-fidelity, ultra-high-resolution videos, addressing the critical need for scalable video generation in applications like film production and virtual reality.
PixelWizard addresses the coupled bottleneck of optimization instability and prohibitive computational costs in high-resolution video generation by hierarchically decoupling global structure from fine-grained detail, achieving over 10x acceleration for native 2K/4K video generation with superior visual quality.
High-resolution video generation faces a coupled bottleneck of optimization instability and prohibitive computational costs. The massive expansion of the token sequence not only biases optimization toward local textures at the expense of global coherence, leading to structural collapse, but also imposes prohibitive training costs and severe inference latency. To address this, we propose PixelWizard, a framework that hierarchically decouples global structure modeling from fine-grained detail synthesis. PixelWizard first establishes a compact spatiotemporal anchor to concentrate dense structural priors, which then guides fine-grained generation at high resolution. This mitigates the local optimization bias to ensure structural stability without compromising high-frequency details. Leveraging this structural stability, we introduce Noise-Span Aligned Shortcut Training to break the inference bottleneck. By explicitly modeling the step size, this mechanism allows the model to traverse the generation trajectory with large steps. Crucially, we incorporate Exponential Index-Biased Sampling and Adaptive Noise-Span Calibration to align optimization with the shifted noise schedules of high-resolution grids, ensuring robust few-step inference without incurring the heavy overhead of distillation. Extensive experiments demonstrate that PixelWizard achieves superior visual quality while accelerating the generative sampling of native 2K/4K videos by over 10x.