Video stylization typically transforms entire scenes, while local artistic editing requires precise region--style association, spatial control, and temporal consistency. We introduce LociArt, a unified paradigm for free-style local video stylization that explicitly specifies where to stylize and what style to apply while preserving non-target content. At its core, GRNStyle builds on Generative Refinement Networks (GRN) and introduces a polar style-aware mask that embeds style identities as equal-norm directions within selected regions. Complementary representation--geometry conditioning combines encoded mask features with raw-mask spatial support, integrating spatiotemporal context with geometric control. To provide paired supervision, we develop an automated pipeline combining global stylization, video object removal, matting, and compositing, and construct the LociArt-100K Dataset with 100K local stylization pairs. We further establish the LociArt Benchmark, comprising 80 source videos and 84 target masks covering 168 foreground/background stylization cases, with a Style Prototype Similarity (SPS) evaluation scheme for regional style quality and complementary measures of non-target preservation and temporal consistency. Experiments show that GRNStyle outperforms existing methods in regional style quality and spatial control while maintaining strong temporal consistency. Qualitative examples further illustrate joint stylization of multiple regions with the same or different learned styles, despite training only on single-region, single-style pairs. Together, these components provide a unified foundation for learning and evaluating controllable local video stylization. Project page: https://higherhu.github.io/lociart/.
Figures & tables
Figure 1: Overview of the data construction pipeline. Source and object-removed videos are globally stylized and filtered, then combined through mask-guided matting to produce foreground- and background-stylized pairs for human verification.
Figure 2: Overview of GRNStyle. (a) Polar style-aware masks bind styles to regions and support multi-region composition. (b) Anchored mask evidence and geometric support are combined with source features through block-wise gating to guide local video stylization.
Method
Venue
SPS-fg ↑
SPS-bg ↑
CLIPT-fg ↑
CLIPT-bg ↑
PSNR-fg ↑
PSNR-bg ↑
TC ↑
Style ↑
Preserve ↑
LEAST ∗†‡ [ 35 ]
arXiv ’24
0.283
0.302
0.158
0.173
-
-
0.953
2.20
4.50
VACE-1.3B [ 14 ]
ICCV ’25
0.156
0.191
0.114
0.132
32.35
29.17
0.981
0.05
5.00
UniVST ∗‡ [ 10 ]
TPAMI ’25
0.191
0.286
0.127
0.165
27.69
23.92
0.969
1.83
4.40
RegionRoute ‡ [ 12 ]
CVPR ’26
0.317
0.491
0.154
0.183
29.38
13.17
0.955
2.41
2.75
StyleGallery ∗†‡ [ 11 ]
CVPR ’26
0.230
0.325
0.146
0.183
-
-
0.957
3.39
4.77
CoinVE-Edit [ 13 ]
arXiv ’26
0.248
0.251
0.153
0.159
20.71
16.76
0.983
3.11
3.93
Table 1: Comparison on the LociArt Benchmark. The fg/bg suffixes denote stylization of the target mask M and its complement 1−M . FT denotes local-stylization fine-tuning, and ∗ denotes training-free methods. † marks direct mask-compositing methods, for which PSNR is omitted because non-edited regions are copied from the source. ‡ marks image-based methods. Best and second-best results are shown in bold and underlined, respectively.
Figure 3: Qualitative comparison of local video stylization. Examples evaluate stylization inside the editing region, preservation outside it, and temporal consistency. Foreground and background cases highlight leakage, distortion of preserved content, and boundary inconsistency. GRNStyle produces localized styles with coherent boundaries and stable non-target content.
Table 3: Replay-free style expansion from 10 to 15 classes. Expanded models are trained only on the five new classes. Retention and acquisition are evaluated on 1,680 original-style and 840 new-style cases, respectively.
Figure 6: Data-construction ablation. For curly hair (a) and head contours (b), columns compare direct compositing, removal without matting, and the full pipeline. Rows show background- and foreground-stylized targets; insets highlight residual silhouettes, clipped contours, and boundary quality.
Appendix figures & tables5 assets
Supplementary material from the paper’s appendix.
Appendix
Figure 7: Representative rejection cases from the data pipeline.
Figure 8: Statistics of the LociArt-100K Dataset. The 100K qualified pairs cover 18 style categories. Area ratios refer to foreground object masks Mobj , not the complementary background editing masks.
Figure 9: SPS prototype visualization and validation. (a) t-SNE visualization of 80 CSD prototypes across 18 styles. (b) Row-normalized confusion matrix under leave-one-source-out validation, achieving 93.61% macro-averaged accuracy.
Style category
Canonical phrase
in bench10
Children’s Book
children’s book illustration
✓
Claymation
claymation
✓
Comic Book
comic book
Disney
Disney animation
Ghibli
Studio Ghibli anime
✓
Japanese Anime
Japanese anime
Appendix
Table 4: Canonical style phrases for regional CLIPT. A check mark identifies the ten-style comparison subset.
Hyperparameter
Setting
Appendix
Table 5: Main training configuration for GRNStyle-2B.
While image stylization has been studied extensively, video stylization remains a critical and largely unsolved challenge in the field of intelligent content creation. Existing methods, usually utilizing a reference image as the style prior, suffer from content leakage, data scarcity and limited adaptability to long videos, leading to suboptimal results with severe style drift and motion distortion. For these issues, we present EchoStyle, a scalable text-driven framework to achieve high-quality stylization of videos with arbitrary lengths. To start with, we construct a video-to-video architecture to appropriately re-fuse the video content and the text style. To address data scarcity, we pioneer an automatic reverse-synthesis pipeline to establish V-Style20k, a large-scale stylization dataset of 20k high-quality video pairs. To facilitate long video stylization, we devise an init-follow-mode mechanism along with a sliding-window inference strategy. Extensive experiments demonstrate EchoStyle's excellent performance across a wide range of artistic styles, even comparable to leading closed-source solutions.
Huaqiu Li, Jiahao Wang, Sijia Cai +4
The Hong Kong University of Science and Technology · Alibaba Group · Xi’an Jiaotong University
Image stylization with latent-diffusion models entangles two independently refined axes: what a region depicts and how it is depicted. Such pipelines expose only global controls, yet professional retouching demands deliberate, region-specific control. We lift two conditioning weights already present in a ControlNet + IP-Adapter stylization pipeline from global scalars to per-location spatial maps, yielding local, per-axis control of content and style in a single generative pass. Because the two weights act on disjoint pathways, adjusting them independently spans a 2x2 retouching vocabulary, from free regeneration to identity preservation. We validate that edits stay confined to the retouched region and that each weight predominantly steers its own axis. Our approach requires no retraining and drops unchanged into any such pipeline.
Video style transfer aims to render videos in a target artistic style while preserving content, structure, and motion. While image stylization has advanced rapidly, video stylization remains challenging due to temporal inconsistency. Most existing methods stylize frames or keyframes and enforce consistency via heuristic temporal propagation, which is brittle under occlusions, disocclusions, and long-term motion, leading to drift and flickering artifacts. We argue that a fundamental bottleneck lies in the lack of large-scale triplet data and a principled training paradigm that jointly models and disentangles style, content, and motion.To address this, we introduce VISTA-1000, a synthetic dataset with 1,000 styles and motion-aligned triplets of style reference, clean video, and stylized video, and propose a diffusion-transformer-based in-context video style transfer framework with a lightweight style adapter for robust style extraction. Extensive experiments demonstrate SOTA performance in style fidelity, temporal consistency, and content preservation.
Yiren Song, Wangzi Yao, Haofan Wang +1
Show Lab, National University of Singapore · Institute of Automation, Chinese Academy of Sciences · Lovart AI