FasterVAR: Plug-and-Play Acceleration for Visual Autoregressive Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914606185709568 |
|---|---|
| author | Li, Senmao Wang, Kai Khan, Salman Khan, Fahad Shahbaz Yang, Jian Wang, Yaxing |
| author_facet | Li, Senmao Wang, Kai Khan, Salman Khan, Fahad Shahbaz Yang, Jian Wang, Yaxing |
| contents | Visual Autoregressive (VAR) modeling departs from the next-token prediction paradigm of traditional Autoregressive (AR) models through next-scale prediction, enabling high-quality image generation. However, the VAR paradigm suffers from sharply increased computational complexity and running time at large-scale steps. Although existing acceleration methods reduce runtime for large-scale steps, but rely on manual step selection and overlook the varying importance of different stages in the generation process. To address this challenge, we present FasterVAR, a systematic study and plug-and-play acceleration framework for VAR models. Our analysis shows that early steps are critical for preserving semantic and structural consistency and should remain intact,while later steps mainly refine details and can be pruned or approximated for acceleration. Building on these insights, FasterVAR introduces a plug-and-play acceleration strategy that exploits semantic irrelevance and low-rank properties in late-stage computations, without requiring additional training. Our proposed FasterVAR achieves up to 3.4x speedup with almost no performance loss. consistently outperforming existing acceleration baselines.These results highlight stage-aware design as a powerful principle for efficient visual autoregressive image generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_16483 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | FasterVAR: Plug-and-Play Acceleration for Visual Autoregressive Models Li, Senmao Wang, Kai Khan, Salman Khan, Fahad Shahbaz Yang, Jian Wang, Yaxing Computer Vision and Pattern Recognition Visual Autoregressive (VAR) modeling departs from the next-token prediction paradigm of traditional Autoregressive (AR) models through next-scale prediction, enabling high-quality image generation. However, the VAR paradigm suffers from sharply increased computational complexity and running time at large-scale steps. Although existing acceleration methods reduce runtime for large-scale steps, but rely on manual step selection and overlook the varying importance of different stages in the generation process. To address this challenge, we present FasterVAR, a systematic study and plug-and-play acceleration framework for VAR models. Our analysis shows that early steps are critical for preserving semantic and structural consistency and should remain intact,while later steps mainly refine details and can be pruned or approximated for acceleration. Building on these insights, FasterVAR introduces a plug-and-play acceleration strategy that exploits semantic irrelevance and low-rank properties in late-stage computations, without requiring additional training. Our proposed FasterVAR achieves up to 3.4x speedup with almost no performance loss. consistently outperforming existing acceleration baselines.These results highlight stage-aware design as a powerful principle for efficient visual autoregressive image generation. |
| title | FasterVAR: Plug-and-Play Acceleration for Visual Autoregressive Models |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.16483 |