pith. machine review for the scientific record. sign in

arxiv: 2604.08646 · v1 · submitted 2026-04-09 · 💻 cs.CV

Recognition: unknown

InsEdit: Towards Instruction-based Visual Editing via Data-Efficient Video Diffusion Models Adaptation

Zhefan Rao , Bin Zou , Haoxuan Che , Xuanhua He , Chong Hou Choi , Yanheng Li , Rui Liu , Qifeng Chen
Authors on Pith no claims yet

Pith reviewed 2026-05-10 17:35 UTC · model grok-4.3

classification 💻 cs.CV
keywords instruction-based video editingvideo diffusion modelsdata-efficient adaptationMutual Context Attentionvideo editing benchmarksimage editingvisual editing architecture
0
0 comments X

The pith

A video generation backbone can become a strong instruction-based editor with only around 100,000 edited video clips.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that adapting a pre-trained video diffusion model does not demand enormous amounts of specialized editing data. Instead, a Mutual Context Attention pipeline generates aligned video pairs that let edits begin at any point in a clip. This data-efficient recipe, combined with image editing examples, produces state-of-the-art open-source results on video instruction benchmarks while also enabling image edits without further changes.

Core claim

InsEdit adapts HunyuanVideo-1.5 into an instruction-based editor by pairing a visual editing architecture with a Mutual Context Attention data pipeline. The pipeline constructs aligned video pairs where text-driven edits can start mid-clip rather than only from the first frame. Trained on O(100)K video editing examples plus image data, the model reaches state-of-the-art performance among open-source methods on video instruction editing benchmarks and supports unmodified image editing.

What carries the argument

Mutual Context Attention (MCA) pipeline, which produces aligned video pairs from existing data so edits can occur at arbitrary points within a clip.

If this is right

  • Video instruction editing becomes practical for open-source developers with modest data budgets.
  • The same adapted model handles both video and image editing without separate training branches.
  • Edits can be applied starting at any frame rather than being restricted to the beginning of a sequence.
  • Data pipelines that create mid-clip edit pairs reduce reliance on manually annotated large-scale video datasets.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Similar alignment techniques might let other diffusion-based generators adapt to editing tasks in domains such as 3D or audio.
  • If MCA scales to even smaller data regimes, zero-shot or few-shot instruction editing could become feasible.
  • The approach suggests that generation models can be repurposed for control tasks more generally whenever paired data can be synthesized through attention mechanisms.

Load-bearing premise

The Mutual Context Attention pipeline produces video pairs that are aligned and high-quality enough for effective model adaptation without large-scale editing data.

What would settle it

A controlled experiment showing that models trained on MCA-generated pairs perform no better than those trained on randomly paired or temporally misaligned clips when evaluated on the same video instruction benchmarks.

Figures

Figures reproduced from arXiv: 2604.08646 by Bin Zou, Chong Hou Choi, Haoxuan Che, Qifeng Chen, Rui Liu, Xuanhua He, Yanheng Li, Zhefan Rao.

Figure 1
Figure 1. Figure 1: Representative visual editing results of InsEdit. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Training efficiency of adapting video generation [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overview of the InsEdit architecture. The semantic [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Automatic pipeline for video editing data construction. Starting from keyword prompts, the pipeline synthesizes [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mutual Context Attention (MCA) for paired video generation. Two denoising branches interact within a shared DiT [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Examples from the MCA-based data construction [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Qualitative results on video instruction editing. Representative examples show that InsEdit follows diverse editing [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Qualitative comparison with some baselines on [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Additional qualitative results on GEdit. Supplementary examples illustrate the image editing performance of InsEdit [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Additional qualitative video editing results. More supplementary examples illustrate the editing performance of [PITH_FULL_IMAGE:figures/full_fig_p013_10.png] view at source ↗
read the original abstract

Instruction-based video editing is a natural way to control video content with text, but adapting a video generation model into an editor usually appears data-hungry. At the same time, high-quality video editing data remains scarce. In this paper, we show that a video generation backbone can become a strong video editor without large scale video editing data. We present InsEdit, an instruction-based editing model built on HunyuanVideo-1.5. InsEdit combines a visual editing architecture with a video data pipeline based on Mutual Context Attention (MCA), which creates aligned video pairs where edits can begin in the middle of a clip rather than only from the first frame. With only O(100)K video editing data, InsEdit achieves state-of-the-art results among open-source methods on our video instruction editing benchmarks. In addition, because our training recipe also includes image editing data, the final model supports image editing without any modification.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The paper introduces InsEdit, an instruction-based video editing model adapted from the HunyuanVideo-1.5 diffusion backbone. It uses a Mutual Context Attention (MCA) pipeline to synthesize O(100)K temporally coherent, instruction-aligned video editing pairs (with edits possible from arbitrary starting frames), enabling data-efficient fine-tuning. The resulting model reports state-of-the-art performance among open-source methods on the authors' video instruction editing benchmarks and, via mixed image-editing data in training, supports unmodified image editing.

Significance. If the benchmark results and ablation studies hold, the work demonstrates that high-quality instruction-based video editing is achievable with far less curated data than previously assumed, by leveraging a targeted data-generation pipeline rather than scale. This has clear practical value for lowering the barrier to specialized editing models. The MCA construction and the observed video-to-image generalization are concrete contributions that could influence future data-efficient adaptation strategies in diffusion-based generation.

minor comments (2)
  1. [§3.2] §3.2 and Figure 3: the description of how MCA enforces temporal consistency across the generated pairs would benefit from an explicit statement of the attention mask construction (e.g., whether it is strictly causal or allows bidirectional context within the edit window).
  2. [Table 2] Table 2: the quantitative comparison table would be strengthened by reporting the exact number of training samples used for each baseline method, to make the data-efficiency claim directly comparable.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive review and the recommendation to accept the paper. We are pleased that the data-efficient adaptation via Mutual Context Attention and the video-to-image generalization are viewed as having practical value for future work.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper describes an empirical adaptation of a video diffusion backbone (HunyuanVideo-1.5) using a Mutual Context Attention (MCA) data pipeline to synthesize O(100K) instruction-aligned editing pairs, followed by standard fine-tuning and benchmarking against open-source baselines. All load-bearing elements—the MCA pair construction, mixed image/video training recipe, and reported SOTA performance—are presented as directly implemented and measured quantities with explicit construction details and external comparisons in the manuscript. No equations, predictions, or uniqueness claims reduce by definition or self-citation to the inputs themselves; the central result is an observed empirical outcome rather than a self-referential fit or renamed ansatz.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 1 invented entities

Populated from abstract only; no explicit free parameters or detailed axioms are stated. MCA is treated as the key invented component for the data pipeline.

axioms (1)
  • domain assumption A pre-trained video diffusion model can be adapted into an instruction-based editor through architectural additions and targeted data alignment.
    Implicit foundation for building InsEdit on HunyuanVideo-1.5.
invented entities (1)
  • Mutual Context Attention (MCA) no independent evidence
    purpose: Creates aligned video pairs allowing edits to begin mid-clip for efficient training data generation.
    Introduced as the core of the video data pipeline.

pith-pipeline@v0.9.0 · 5481 in / 1311 out tokens · 94319 ms · 2026-05-10T17:35:34.368402+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 42 canonical work pages · 16 internal anchors

  1. [1]

    Qingyan Bai, Qiuyu Wang, Hao Ouyang, Yue Yu, Hanlin Wang, Wen Wang, Ka Leong Cheng, Shuailei Ma, Yanhong Zeng, Zichen Liu, Yinghao Xu, Yujun Shen, and Qifeng Chen. 2025. Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset.CoRRabs/2510.15742 (2025). doi:10.48550/arXiv. 2510.15742

    work page internal anchor Pith review doi:10.48550/arxiv 2025

  2. [2]

    Stephen Batifol, Alexander Lorenz, Ajay Jain, Sander Becker, Tom Bos, Simon Buchholz, Itai Caspi, Eyal Cohen, Shuyang Ge, Xuan Li, et al . 2025. FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space.arXiv preprint arXiv:2506.15742(2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  3. [3]

    Tim Brooks, Aleksander Holynski, and Alexei A. Efros. 2023. InstructPix2Pix: Learning to Follow Image Editing Instructions. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 18392–18402. doi:10.1109/ CVPR52729.2023.01764

    work page arXiv 2023

  4. [4]

    Mingdeng Cao, Xintao Wang, Zhongang Qi, Ying Shan, Xiaohu Qie, and Yinqiang Zheng. 2023. MasaCtrl: Tuning-Free Mutual Self-Attention Control for Consistent Image Synthesis and Editing. arXiv:2304.08465 [cs.CV] https://arxiv.org/abs/ 2304.08465

    work page arXiv 2023

  5. [5]

    Junyi Chen, Tong He, Zhoujie Fu, Pengfei Wan, Kun Gai, and Weicai Ye. 2026. VINO: A Unified Visual Generator with Interleaved Omnimodal Context.arXiv preprint arXiv:2601.02358(2026)

    work page arXiv 2026

  6. [6]

    Yuren Cong, Mengmeng Xu, Christian Simon, Shoufa Chen, Jiawei Ren, Yanping Xie, Juan-Manuel Perez-Rua, Bodo Rosenhahn, Tao Xiang, and Sen He. 2023. Flatten: optical flow-guided attention for consistent text-to-video editing.arXiv preprint arXiv:2310.05922(2023)

    work page arXiv 2023

  7. [7]

    Tri Dao. 2023. FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning. arXiv:2307.08691 [cs.LG]

    work page internal anchor Pith review arXiv 2023

  8. [8]

    Chaorui Deng, Deyao Zhu, Kunchang Li, Chenhui Gou, Feng Li, Zeyu Wang, Shu Zhong, Weihao Yu, Xiaonan Nie, Ziang Song, and Haoqi Fan. 2025. Emerging Properties in Unified Multimodal Pretraining.arXiv preprint arXiv:2505.14683 (2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  9. [9]

    Michal Geyer, Omer Bar-Tal, Shai Bagon, and Tali Dekel. 2023. Tokenflow: Consis- tent diffusion features for consistent video editing.arXiv preprint arXiv:2307.10373 (2023)

    work page arXiv 2023

  10. [10]

    Haoyang He, Jie Wang, Jiangning Zhang, Zhucun Xue, Xingyuan Bu, Qiangpeng Yang, Shilei Wen, and Lei Xie. 2025. OpenVE-3M: A Large-Scale High-Quality Dataset for Instruction-Guided Video Editing.arXiv preprint arXiv:2512.07826 (2025)

    work page arXiv 2025

  11. [11]

    Amir Hertz, Ron Mokady, Jay Tenenbaum, Kfir Aberman, Yael Pritch, and Daniel Cohen-Or. 2022. Prompt-to-Prompt Image Editing with Cross Attention Control. arXiv:2208.01626 [cs.CV] https://arxiv.org/abs/2208.01626

    work page internal anchor Pith review arXiv 2022

  12. [12]

    Wenyi Hong, Ming Ding, Wendi Zheng, Xinghan Liu, and Jie Tang. 2022. Cogvideo: Large-scale pretraining for text-to-video generation via transformers. arXiv preprint arXiv:2205.15868(2022)

    work page internal anchor Pith review arXiv 2022

  13. [13]

    Zeyinzi Jiang, Zhen Han, Chaojie Mao, Jingfeng Zhang, Yulin Pan, and Yu Liu

  14. [14]

    Vace: All-in-one video creation and editing.arXiv preprint arXiv:2503.07598 (2025)

    work page arXiv 2025

  15. [15]

    Xuan Ju, Tianyu Wang, Yuqian Zhou, He Zhang, Qing Liu, Nanxuan Zhao, Zhifei Zhang, Yijun Li, Yuanhao Cai, Shaoteng Liu, Daniil Pakhomov, Zhe Lin, Soo Ye Kim, and Qiang Xu. 2025. EditVerse: Unifying Image and Video Editing and Generation with In-Context Learning.arXiv preprint arXiv:2509.20360(2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  16. [16]

    Ozgur Kara, Bariscan Kurtkaya, Hidir Yesiltepe, James M Rehg, and Pinar Ya- nardag. 2024. Rave: Randomized noise shuffling for fast and consistent video editing with diffusion models. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 6507–6516

    2024

  17. [17]

    Weijie Kong, Qi Tian, Zijian Zhang, Rox Min, Zuozhuo Dai, Jin Zhou, Jiangfeng Xiong, Xin Li, Bo Wu, Jianwei Zhang, et al. 2024. Hunyuanvideo: A systematic framework for large video generative models.arXiv preprint arXiv:2412.03603 (2024)

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  18. [18]

    Max Ku, Cong Wei, Weiming Ren, Harry Yang, and Wenhu Chen. 2024. Anyv2v: A tuning-free framework for any video-to-video editing tasks.arXiv preprint arXiv:2403.14468(2024)

    work page arXiv 2024

  19. [19]

    Maksim Kuprashevich, Grigorii Alekseenko, Irina Tolstykh, Georgii Fedorov, Bulat Suleimanov, Vladimir Dokholyan, and Aleksandr Gordeev. 2025. Nohu- mansrequired: Autonomous high-quality image editing triplet mining.arXiv preprint arXiv:2507.14119(2025)

    work page arXiv 2025

  20. [20]

    Sen Liang, Zhentao Yu, Zhengguang Zhou, Teng Hu, Hongmei Wang, Yi Chen, Qin Lin, Yuan Zhou, Xin Li, Qinglin Lu, et al. 2025. OmniV2V: Versatile Video Generation and Editing via Dynamic Content Manipulation.arXiv preprint arXiv:2506.01801(2025)

    work page arXiv 2025

  21. [21]

    Xinyao Liao, Xianfang Zeng, Ziye Song, Zhoujie Fu, Gang Yu, and Guosheng Lin. 2025. In-Context Learning with Unpaired Clips for Instruction-based Video Editing.CoRRabs/2510.14648 (2025). doi:10.48550/arXiv.2510.14648

    work page doi:10.48550/arxiv.2510.14648 2025

  22. [22]

    Bin Lin, Zongjian Li, Xinhua Cheng, Yuwei Niu, Yang Ye, Xianyi He, Shenghai Yuan, Wangbo Yu, Shaodong Wang, Yunyang Ge, et al. 2025. UniWorld: High- Resolution Semantic Encoders for Unified Visual Understanding and Generation. arXiv preprint arXiv:2506.03147(2025)

    work page internal anchor Pith review arXiv 2025

  23. [23]

    Yiqi Lin, Guoqiang Liang, Ziyun Zeng, Zechen Bai, Yanzhe Chen, and Mike Zheng Shou. 2026. Kiwi-Edit: Versatile Video Editing via Instruction and Reference Guidance.CoRRabs/2603.02175 (2026). doi:10.48550/arXiv.2603.02175

    work page internal anchor Pith review doi:10.48550/arxiv.2603.02175 2026

  24. [24]

    Shiyu Liu, Yucheng Han, Peng Xing, Fukun Yin, Rui Wang, Wei Cheng, Jiaqi Liao, Yingming Wang, Honghao Fu, Chunrui Han, Guopeng Li, Yuang Peng, Quan Sun, Jingwei Wu, Yan Cai, Zheng Ge, Ranchen Ming, Lei Xia, Xianfang Zeng, Yibo Zhu, Binxing Jiao, Xiangyu Zhang, Gang Yu, and Daxin Jiang. 2025. Step1X-Edit: A Practical Framework for General Image Editing.arX...

    work page internal anchor Pith review arXiv 2025

  25. [25]

    Jiabin Luo, Junhui Lin, Zeyu Zhang, Biao Wu, Meng Fang, Ling Chen, and Hao Tang. 2025. Univid: The open-source unified video model.arXiv preprint arXiv:2509.24200(2025)

    work page arXiv 2025

  26. [26]

    Xin Luo, Jiahao Wang, Chenyuan Wu, Shitao Xiao, Xiyan Jiang, Defu Lian, Jiajun Zhang, Dong Liu, et al. 2025. Editscore: Unlocking online rl for image editing via high-fidelity reward modeling.arXiv preprint arXiv:2509.23909(2025)

    work page arXiv 2025

  27. [27]

    Chong Mou, Qichao Sun, Yanze Wu, Pengze Zhang, Xinghui Li, Fulong Ye, Song- tao Zhao, and Qian He. 2025. Instructx: Towards unified visual editing with mllm guidance.arXiv preprint arXiv:2510.08485(2025)

    work page arXiv 2025

  28. [28]

    Chenyang Qi, Xiaodong Cun, Yong Zhang, Chenyang Lei, Xintao Wang, Ying Shan, and Qifeng Chen. 2023. Fatezero: Fusing attentions for zero-shot text-based video editing. InProceedings of the IEEE/CVF International Conference on Computer Vision. 15932–15942

    2023

  29. [29]

    Bosheng Qin, Juncheng Li, Siliang Tang, Tat-Seng Chua, and Yueting Zhuang

  30. [30]

    Instructvid2vid: Controllable video editing with natural language instruc- tions. InICME

  31. [31]

    Zhiyu Tan, Hao Yang, Luozheng Qin, Jia Gong, Mengping Yang, and Hao Li. 2025. Omni-Video: Democratizing Unified Video Understanding and Generation.arXiv preprint arXiv:2507.06119(2025)

    work page arXiv 2025

  32. [32]

    DecartAI Team. 2025. Lucy Edit: Open-Weight Text-Guided Video Edit- ing. https://d2drjpuinn46lb.cloudfront.net/Lucy_Edit__High_Fidelity_Text_ Guided_Video_Editing.pdf

    2025

  33. [33]

    Kling Team. 2025. Kling-Omni Technical Report. arXiv:2512.16776 [cs.CV]

    work page arXiv 2025

  34. [34]

    Tencent Hunyuan Foundation Model Team. 2025. HunyuanVideo 1.5 Technical Report. arXiv:2511.18870 [cs.CV] https://arxiv.org/abs/2511.18870

    work page internal anchor Pith review arXiv 2025

  35. [35]

    Team Wan, Ang Wang, Baole Ai, Bin Wen, Chaojie Mao, Chen-Wei Xie, Di Chen, Feiwu Yu, Haiming Zhao, Jianxiao Yang, et al. 2025. Wan: Open and advanced large-scale video generative models.arXiv preprint arXiv:2503.20314(2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  36. [36]

    Yuhan Wang, Siwei Yang, Bingchen Zhao, Letian Zhang, Qing Liu, Yuyin Zhou, and Cihang Xie. 2025. Gpt-image-edit-1.5 m: A million-scale, gpt-generated image dataset.arXiv preprint arXiv:2507.21033(2025)

    work page arXiv 2025

  37. [37]

    Cong Wei, Quande Liu, Zixuan Ye, Qiulin Wang, Xintao Wang, Pengfei Wan, Kun Gai, and Wenhu Chen. 2025. Univideo: Unified understanding, generation, and editing for videos.arXiv preprint arXiv:2510.08377(2025)

    work page arXiv 2025

  38. [38]

    Chenfei Wu, Jiahao Li, Jingren Zhou, Junyang Lin, Kaiyuan Gao, Kun Yan, Sheng ming Yin, Shuai Bai, Xiao Xu, Yilei Chen, Yuxiang Chen, Zecheng Tang, Zekai Zhang, Zhengyi Wang, An Yang, Bowen Yu, Chen Cheng, Dayiheng Liu, Deqing Li, Hang Zhang, Hao Meng, Hu Wei, Jingyuan Ni, Kai Chen, Kuan Cao, Liang Peng, Lin Qu, Minggang Wu, Peng Wang, Shuting Yu, Tingkun...

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  39. [39]

    Jay Zhangjie Wu, Yixiao Ge, Xintao Wang, Stan Weixian Lei, Yuchao Gu, Yufei Shi, Wynne Hsu, Ying Shan, Xiaohu Qie, and Mike Zheng Shou. 2023. Tune-a- video: One-shot tuning of image diffusion models for text-to-video generation. InProceedings of the IEEE/CVF International Conference on Computer Vision. 7623– 7633

    2023

  40. [40]

    Xiaoshi Wu, Yixuan Jiao, Wen Wang, Zhiyu Tan, Xialei Lyu, Hanyu Li, Shijie Guo, Zijian Zhang, Xin Zhang, Ji Zhu, et al . 2025. OmniGen2: Exploration to Advanced Multimodal Generation.arXiv preprint arXiv:2506.18871(2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  41. [41]

    Yuhui Wu, Liyi Chen, Ruibin Li, Shihao Wang, Chenxi Xie, and Lei Zhang. 2025. InsViE-1M: Effective Instruction-based Video Editing with Elaborate Dataset Construction.CoRRabs/2503.20287 (2025). doi:10.48550/arXiv.2503.20287

    work page doi:10.48550/arxiv.2503.20287 2025

  42. [42]

    Bin Xia, Jiyang Liu, Yuechen Zhang, Bohao Peng, Ruihang Chu, Yitong Wang, Xinglong Wu, Bei Yu, and Jiaya Jia. 2025. DreamVE: Unified Instruction-based Image and Video Editing.CoRRabs/2508.06080 (2025). doi:10.48550/arXiv.2508. 06080

    work page doi:10.48550/arxiv.2508 2025

  43. [43]

    Bin Xia, Bohao Peng, Yuechen Zhang, Junjia Huang, Jiyang Liu, Jingyao Li, Haoru Tan, Sitong Wu, Chengyao Wang, Yitong Wang, Xinglong Wu, Bei Yu, and Jiaya Jia. 2025. DreamOmni2: Multimodal Instruction-based Editing and Generation. arXiv:2510.06679 [cs.CV] https://arxiv.org/abs/2510.06679

    work page arXiv 2025

  44. [44]

    Xiangpeng Yang, Linchao Zhu, Hehe Fan, and Yi Yang. 2025. VideoGrain: Modu- lating space-time attention for multi-grained video editing. InICLR

    2025

  45. [45]

    Yang Ye, Xianyi He, Zongjian Li, Bin Lin, Shenghai Yuan, Zhiyuan Yan, Bohan Hou, and Li Yuan. 2025. Imgedit: A unified image editing dataset and benchmark. arXiv preprint arXiv:2505.20275(2025)

    work page internal anchor Pith review arXiv 2025

  46. [46]

    Zixuan Ye, Xuanhua He, Quande Liu, Qiulin Wang, Xintao Wang, Pengfei Wan, Di Zhang, Kun Gai, Qifeng Chen, and Wenhan Luo. 2025. UNIC: Unified In-Context Video Editing. arXiv:2506.04216 [cs.CV] 9

    work page arXiv 2025

  47. [47]

    Shoubin Yu, Difan Liu, Ziqiao Ma, Yicong Hong, Yang Zhou, Hao Tan, Joyce Chai, and Mohit Bansal. 2025. VEGGIE: Instructional Editing and Reasoning Video Concepts with Grounded Generation.arXiv preprint arXiv:2503.14350(2025)

    work page arXiv 2025

  48. [48]

    Zhongwei Zhang, Fuchen Long, Wei Li, Zhaofan Qiu, Wu Liu, Ting Yao, and Tao Mei. 2025. Region-Constraint In-Context Generation for Instructional Video Editing.arXiv preprint arXiv:2512.17650(2025)

    work page arXiv 2025

  49. [49]

    Hanshen Zhu, Zhen Zhu, Kaile Zhang, Yiming Gong, Yuliang Liu, and Xi- ang Bai. 2025. Training-free Geometric Image Editing on Diffusion Models. arXiv:2507.23300 [cs.CV] https://arxiv.org/abs/2507.23300

    work page arXiv 2025

  50. [50]

    Bojia Zi, Penghui Ruan, Marco Chen, Xianbiao Qi, Shaozhe Hao, Shihao Zhao, Youze Huang, Bin Liang, Rong Xiao, and Kam-Fai Wong. 2025. Se\˜ norita-2M: A High-Quality Instruction-based Dataset for General Video Editing by Video Specialists.arXiv preprint arXiv:2502.06734(2025). 10 A MCA Schedule Details In the main paper, we describe MCA as a general framew...

    work page arXiv 2025