通过三重压缩提升1080p视频生成速度，快76倍且画质损失小。
HiStream: Efficient High-Resolution Video Generation via Redundancy-Eliminated Streaming
- 分三步压缩：低分辨率去噪、分块缓存、减少后续步骤。
- 相比基线模型提速76.2倍，画质几乎无损，1080p上表现最佳。
- 适合需要高效生成高清视频的影视与数字媒体从业者。
高分辨率视频生成对数字媒体和电影至关重要，但扩散模型的二次复杂度导致计算瓶颈，使实际推理难以实现。为此，我们提出HiStream，一种高效的自回归框架，系统性地在三个维度上消除冗余：(i) 空间压缩：先在低分辨率去噪，再用缓存特征在高分辨率精修；(ii) 时间压缩：采用分块策略搭配固定大小锚点缓存，保证推理速度稳定；(iii) 时间步压缩：对后续缓存条件化的块应用更少去噪步骤。在1080p基准测试中，主模型（i+ii）达到最先进的视觉质量，同时相比Wan2.1基线实现最高76.2倍的去噪加速，且质量损失可忽略。其更快版本HiStream+（i+ii+iii）相较基线提速达107.5倍，实现了速度与质量的出色平衡，使高分辨率视频生成变得实用且可扩展。
原文摘要 · Abstract (English)
High-resolution video generation, while crucial for digital media and film, is computationally bottlenecked by the quadratic complexity of diffusion models, making practical inference infeasible. To address this, we introduce HiStream, an efficient autoregressive framework that systematically reduces redundancy across three axes: i) Spatial Compression: denoising at low resolution before refining at high resolution with cached features; ii) Temporal Compression: a chunk-by-chunk strategy with a fixed-size anchor cache, ensuring stable inference speed; and iii) Timestep Compression: applying fewer denoising steps to subsequent, cache-conditioned chunks. On 1080p benchmarks, our primary HiStream model (i+ii) achieves state-of-the-art visual quality while demonstrating up to 76.2x faster denoising compared to the Wan2.1 baseline and negligible quality loss. Our faster variant, HiStream+, applies all three optimizations (i+ii+iii), achieving a 107.5x acceleration over the baseline, offering a compelling trade-off between speed and quality, thereby making high-resolution video generation both practical and scalable.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。