CVAIJun 2, 2025

Playing with Transformer at 30+ FPS via Next-Frame Diffusion

arXiv:2506.01380v214 citationsh-index: 12
Originality Highly original
AI Analysis

This enables interactive video content creation and streaming applications, representing a strong specific gain in video generation technology.

The paper tackles the challenge of real-time video generation by proposing Next-Frame Diffusion (NFD), an autoregressive diffusion transformer that achieves over 30 FPS on an A100 GPU with a 310M model, beating baselines in visual quality and sampling efficiency.

Autoregressive video models offer distinct advantages over bidirectional diffusion models in creating interactive video content and supporting streaming applications with arbitrary duration. In this work, we present Next-Frame Diffusion (NFD), an autoregressive diffusion transformer that incorporates block-wise causal attention, enabling iterative sampling and efficient inference via parallel token generation within each frame. Nonetheless, achieving real-time video generation remains a significant challenge for such models, primarily due to the high computational cost associated with diffusion sampling and the hardware inefficiencies inherent to autoregressive generation. To address this, we introduce two innovations: (1) We extend consistency distillation to the video domain and adapt it specifically for video models, enabling efficient inference with few sampling steps; (2) To fully leverage parallel computation, motivated by the observation that adjacent frames often share the identical action input, we propose speculative sampling. In this approach, the model generates next few frames using current action input, and discard speculatively generated frames if the input action differs. Experiments on a large-scale action-conditioned video generation benchmark demonstrate that NFD beats autoregressive baselines in terms of both visual quality and sampling efficiency. We, for the first time, achieves autoregressive video generation at over 30 Frames Per Second (FPS) on an A100 GPU using a 310M model.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes