HiStream: Efficient High-Resolution Video Generation via Redundancy-Eliminated Streaming

TL;DR

HiStream introduces an efficient autoregressive framework that significantly accelerates high-resolution video generation by reducing redundancy across spatial, temporal, and timestep dimensions, enabling practical and scalable high-quality video synthesis.

Contribution

The paper presents a novel streaming-based approach with three key redundancy reduction strategies, achieving unprecedented speedups in high-resolution video generation.

Findings

Up to 76.2x faster denoising compared to baseline

Achieves state-of-the-art visual quality on 1080p videos

107.5x acceleration with a trade-off between speed and quality

Abstract

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.