Pith. sign in

REVIEW 2 major objections 5 minor 3 cited by

ToonComposer: Streamlining Cartoon Production with Generative Post-Keyframing

T0 review · 2 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read ToonComposer claims that cartoon inbetweening and colorization can be collapsed into a single generative post-keyframing stage that outperforms two-stage pipelines.

desk verdict Genuinely new post-keyframing formulation with a thoughtful evaluation, but the ToonCrafter baseline setup on PKBench is under-specified and could be unfair; deserves review but needs a rebuttal. read the letter →

arxiv 2508.10881 v1 pith:ZANR65PN submitted 2025-08-14 cs.CV cs.AI

Lingen Li , Guangzhi Wang , Zhaoyang Zhang , Yaowei Li , Xiaoyu Li , Qi Dou , Jinwei Gu , Tianfan Xue
show 1 more author
Ying Shan
This is my paper
classification cs.CVcs.AI
keywords cartoonproductionpost-keyframingsparsesketchinjectionspatiallow-rankadaptervideodiffusiontransformerinbetweeningcolorizationsketch-guidedgeneration
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper proposes replacing the separate cartoon-production stages of inbetweening and colorization with one generative 'post-keyframing' stage: given one colored reference frame and one or more keyframe sketches, the model directly produces the full cartoon video. The authors argue that existing two-stage workflows accumulate errors because sketch-interpolation failures carry into the colorization step, whereas a unified model avoids that risk. To make this work in a modern full-attention video diffusion transformer, they introduce sparse sketch injection and a spatial low-rank adapter that adapts the model's appearance to cartoons while leaving its temporal behavior untouched. If the approach holds, artists would only need to draw a few keyframes and let the model fill in the rest, substantially cutting manual labor while improving quality and motion consistency.

What carries the argument

The post-keyframing stage itself is the organizing object: it unifies inbetweening and colorization into one diffusion generation step. Two mechanisms carry the argument. Sparse sketch injection embeds sketch latents as extra tokens with rotary-position-encoding mappings taken from the corresponding video tokens, plus a trainable position-aware residual that lets users adjust control strength. The Spatial Low-Rank Adapter (SLRA) downsamples token features, applies self-attention only along the spatial dimension within each frame, and adds the resulting residual after the main full-attention module, thereby adapting appearance without modifying temporal attention.

What would settle it

Run the SLRA-adapted model against a variant in which the temporal layers are also fine-tuned, using cartoon clips with exaggerated, non-photorealistic motion (squash-and-stretch, held frames, fast action arcs). If the frozen-temporal-prior version does not match or beat the fully adapted version on motion-quality measures and professional-animator preference, the central claim about preserving the temporal prior fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes a new task formulation: given a colored reference frame $f_1$ and a sketch frame $s_j$ at temporal position $j$, a model $G_\theta$ directly generates the K-frame cartoon video, $\{\hat{f}_k\}_{k=1}^K = G_\theta(f_1, s_j, e_{\text{text}})$. The central technical claim is that this can be achieved by (1) injecting sparse sketch tokens into the latent token sequence with position-aware residual connections, and (2) adapting a Diffusion Transformer video foundation model to the cartoon domain with a Spatial Low-Rank Adapter (SLRA), which computes a residual from per-frame spatial attention only, so the base model's temporal prior is preserved. The paper r

Load-bearing premise

The method's load-bearing premise is that in a full-attention video diffusion transformer, cartoon style is mostly spatial—so training only a spatial residual leaves the base model's motion behavior valid for cartoons; if cartoon motion is itself domain-specific, the frozen temporal prior is the wrong prior.

Editorial extensions

If this is right

  • The cartoon production pipeline can shrink from keyframing, inbetweening, and colorization to keyframing plus one generative post-keyframing pass.
  • Because the model accepts a variable number of sketches at arbitrary temporal locations, artists can trade control against automation depending on motion complexity.
  • Region-wise control lets artists leave parts of the sketch blank and have the model generate plausible content in those regions, further reducing manual input.
  • With fine-tuning on a small dataset, the same post-keyframing approach transfers to 3D-rendered animation, suggesting a broader production pipeline.
  • If the reported benchmarks hold, unified post-keyframing avoids the cross-stage error accumulation that plagues interpolation-then-colorization pipelines.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The success of SLRA rests on an untested separability claim: if cartoon motion itself is domain-specific, preserving the natural-video temporal prior wholesale could limit stylized timing and exaggeration; a direct test would fine-tune temporal layers as well and compare motion quality.
  • The post-keyframing formulation may transfer to other chained image-to-image production tasks—for example, storyboard-to-preview animation or layout-to-final rendering—where two sequential generative stages are currently used.
  • The adjustable control-strength mechanism suggests an interactive workflow where artists progressively add keyframe sketches until the generated motion matches their intent, effectively using the model as a search tool for motion.
  • Because the benchmark includes human-drawn sketches, the claimed advantage may partly reflect robustness to sketch style variation; a useful extension would measure how performance degrades as sketch style diverges from the training distribution.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper proposes ToonComposer, a DiT-based video generation model that unifies the inbetweening and colorization stages of cartoon production into a single 'post-keyframing' stage. Given one colored reference frame and one or more sparse keyframe sketches at arbitrary temporal positions, it generates a full cartoon video. The method introduces sparse sketch injection (position encoding mapping plus a position-aware residual) and a Spatial Low-Rank Adapter (SLRA) that adapts the Wan 2.1 foundation model to the cartoon domain while keeping its temporal prior intact. Region-wise control is also included to handle blank sketch regions. The authors curate a 37K-clip training dataset (PKData) and a 30-sample benchmark with human-drawn sketches (PKBench), and evaluate against AniDoc, LVCD, and ToonCrafter using reference-based metrics, VBench metrics, and a 47-participant user study, reporting consistent improvements.

Significance. If the reported results hold, ToonComposer is a useful contribution to AI-assisted animation: it directly addresses error accumulation in two-stage pipelines and substantially reduces the manual sketching burden. The introduction of PKBench with human-drawn sketches held out from training is a genuine strength, as is the explicit comparison against assembled two-stage baselines. The SLRA idea, adapting only spatial attention in a full-attention DiT, is interesting and potentially transferable. However, the empirical superiority claim rests on the fairness of the baseline protocol, and the paper does not currently disclose a critical detail about how ToonCrafter is given its required second color keyframe. The SLRA ablation also lacks statistical support for its claimed advantage. These issues are fixable but need to be addressed before the central claims can be accepted.

major comments (2)
  1. [§4.2, Baseline Methods (ToonCrafter protocol)] The comparison against ToonCrafter is underspecified and potentially degenerate. ToonCrafter's sketch-guidance mode requires both a first and a last color frame as input, but the paper states only that dense sketches are interpolated and then ToonCrafter is run, without specifying what is used as the last color frame on PKBench, where only one colored reference frame is provided. If the first reference frame is reused as the last frame, ToonCrafter is forced to generate videos with identical endpoints, which would artificially suppress its motion scores and user preference. Conversely, if the ground-truth last frame is used on the synthetic benchmark, ToonCrafter would receive privileged information. Since Tables 1–3 and the user study are the principal support for the claimed superiority, the exact protocol must be stated. If the first frame was reused, the baseline should be rerun with
  2. [§4.5, Table 4 (SLRA ablation)] The SLRA ablation is the only direct evidence for the load-bearing claim that spatial-only adaptation preserves temporal priors. The margins over the LoRA baseline are small: LPIPS 0.1874 vs 0.1922, DISTS 0.0955 vs 0.1082, CLIP 0.9634 vs 0.9628. No error bars, multiple seeds, or significance tests are reported, and the ablation does not include motion-specific metrics such as VBench motion smoothness or subject consistency, even though the argument is specifically about temporal behavior. With such small differences and no variance estimates, the observed improvements are not clearly distinguishable from noise. The authors should add variance over multiple training runs and ideally evaluate temporal metrics in the ablation.
minor comments (5)
  1. [§4.5, first sentence] "with results detailed in Figure 4" should read "with results detailed in Table 4", since Figure 4 shows the SLRA architecture rather than ablation results.
  2. [References] Several reference IDs appear to be placeholders, e.g., [26] lists arXiv:2503.00000 and [32] lists arXiv:2405.12345. Please verify these. Also, [23] has an author-name typo: "Li Chenn" should likely be "Li Chen".
  3. [§4.4, Human Evaluation] The paper says "randomly select 30 samples from the benchmarks," but PKBench has exactly 30 samples; clarify whether this is the full PKBench set or a subset. In addition, reporting confidence intervals or per-sample variance for the preference rates in Table 3 would help assess the strength of the 70.99%/68.58% figures.
  4. [§4.1, Metrics and Table 1] The metrics list mentions "motion consistency," but Table 1 reports "Motion Smo." and Table 2 reports "M.S." Please unify the terminology and define all abbreviations in the text.
  5. [§3.2, Eq. (4)] The position-aware residual is trained with α=1 and only adjusted at inference. It is unclear whether the model is expected to generalize to other α values without training-time augmentation. A sentence clarifying this would be helpful.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the paper's central claims rest on empirical head-to-head evaluation against external methods, not on a derivation that reduces to its inputs.

full rationale

The load-bearing claim—that the K-frame video in Eq. (1) is generated directly from a colored reference frame and a sparse sketch—is a task definition, not a derived result. The model is trained on PKData and evaluated on held-out benchmarks, with the paper explicitly stating that evaluation movies were excluded from training and that human-drawn sketches are reserved for the real benchmark. No fitted parameter is renamed as a prediction, and no equation is equivalent to its inputs by construction. The SLRA design is supported by an ablation (Table 4) rather than by definitional circularity. The only self-citation of note is [32] (ToonCrafter), which shares author Ying Shan, but it is used as prior work and baseline, not as load-bearing justification for the paper's own quantitative superiority. One methodological ambiguity—the unspecified last color frame for ToonCrafter on PKBench, which only provides one color reference—is a fairness concern for the comparison, but it is not a circularity of the derivation. Overall, the central results are externally evaluated and do not reduce to the model's own assumptions.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central claim rests on the Wan 2.1 prior, the spatial/temporal separability premise, the sufficiency of sparse conditioning, and the sketch-style generalization from synthetic training sketches to human drawings; none of these is derived in the paper, and PKBench is the only held-out check of the last one. No new physical or ontological entities are introduced: 'post-keyframing' is a workflow concept and SLRA is a network module, so the invented-entities list is empty.

free parameters (4)
  • SLRA rank D_low = 144 (LoRA comparison arm uses rank 24 to match parameter count)
    Capacity of the cartoon adaptation branch; set by hand without a reported criterion; the effectiveness claim for SLRA depends on this choice.
  • Sketch control strength alpha = 1 during training, user-adjustable at inference
    Scales the position-aware residual in Eq. (4); it is an inference-time free knob that changes output adherence to sketches.
  • Training schedule = 10 epochs, batch size 16, learning rate 1e-5
    Standard training hyperparameters; the loss (Eq. 12) is fitted to PKData and these settings are not justified by ablations.
  • Mask channel m_in = user-specified binary mask
    Region-wise control adds a per-sample channel (Eq. 10) that the user sets; output content in masked regions is effectively determined by the user input rather than by the model alone.
assumptions (6)
  • domain assumption Wan 2.1 provides a video-generation prior of sufficient quality for cartoon video.
    Adopted as the base model in Sec. 3.1 without justification beyond its public performance.
  • domain assumption Spatial appearance and temporal motion are separable in the full-attention DiT latent space, so a spatial-only residual (SLRA) adapts style while preserving motion.
    Core premise of Sec. 3.3; supported only by the Table 4 ablation, not by analysis of the latent geometry.
  • domain assumption One colored reference frame plus one sketch at an arbitrary temporal location is sufficient conditioning to determine identity, style, and motion of the whole clip.
    The task definition of the post-keyframing stage, Eq. (1) in Sec. 3.1.
  • domain assumption Sketches synthesized by four CNN sketch models plus IC-Sketcher approximate the distribution of real human-drawn sketches.
    Training data construction in Sec. 4.1; PKBench is the held-out test of this premise.
  • standard math Rectified flow velocity-prediction objective trains the conditional model correctly.
    Standard diffusion training result used in Sec. 3.5.
  • domain assumption The inbetweening and colorization stages share an internal correspondence-search mechanism, so merging them does not lose information.
    Motivational claim in Sec. 1 and Sec. 3.1; asserted, not tested, and it justifies the unified stage.

how reviews work

0 comments
Cite this review

Pith. "Pith review of ToonComposer: Streamlining Cartoon Production with Generative Post-Keyframing." pith.science (2026). https://pith.science/paper/ZANR65PN

@misc{pith2026250810881,
  author       = {Pith},
  title        = {Pith review of: ToonComposer: Streamlining Cartoon Production with Generative Post-Keyframing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZANR65PN}},
  note         = {Machine review of arXiv:2508.10881}
}
read the original abstract

Traditional cartoon and anime production involves keyframing, inbetweening, and colorization stages, which require intensive manual effort. Despite recent advances in AI, existing methods often handle these stages separately, leading to error accumulation and artifacts. For instance, inbetweening approaches struggle with large motions, while colorization methods require dense per-frame sketches. To address this, we introduce ToonComposer, a generative model that unifies inbetweening and colorization into a single post-keyframing stage. ToonComposer employs a sparse sketch injection mechanism to provide precise control using keyframe sketches. Additionally, it uses a cartoon adaptation method with the spatial low-rank adapter to tailor a modern video foundation model to the cartoon domain while keeping its temporal prior intact. Requiring as few as a single sketch and a colored reference frame, ToonComposer excels with sparse inputs, while also supporting multiple sketches at any temporal location for more precise motion control. This dual capability reduces manual workload and improves flexibility, empowering artists in real-world scenarios. To evaluate our model, we further created PKBench, a benchmark featuring human-drawn sketches that simulate real-world use cases. Our evaluation demonstrates that ToonComposer outperforms existing methods in visual quality, motion consistency, and production efficiency, offering a superior and more flexible solution for AI-assisted cartoon production.

Figures

Figures reproduced from arXiv: 2508.10881 by the authors.

Figure 1
Figure 1. Video samples generated by ToonComposer using sparse keyframe sketches, featuring scenes from cartoon movies ( [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Comparison between previous cartoon production work [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The model design of ToonComposer. A sparse sketch injection mechanism enables precise control using keyframe sketches, and [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: The structure of the Spatial Low-Rank Adapter (SLRA) [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Examples of different sketch types used during train [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Comparison on the synthetic benchmark among AniDoc, LVCD, ToonCrafter, and our ToonComposer. Zoom-in patches of a [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]
Figure 7
Figure 7. Figure 7: Comparison on the benchmark PKBench, using keyframe sketches drawn by the human artists. Zoom-in patches are shown in the [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Ablation study of the Spatial Low-Rank Adapter [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Illustration of region-wise control in ToonComposer. Without region-wise control, blank areas in keyframe sketches are misin [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: ToonComposer’s flexible controllability with varying keyframe sketches. Using only sketch #1 as the final keyframe and the [PITH_FULL_IMAGE:figures/full_fig_p010_10.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation

    cs.GR 2026-04 unverdicted novelty 7.0 of 10

    MoZoo generates high-fidelity animal videos with fur and muscle dynamics from coarse meshes by extending video diffusion with role-aware RoPE and asymmetric decoupled attention, trained on a new synthetic-to-real dataset.

  2. PeCA: Palette Context Assisted Inference for Test-Time Paint-Bucket Colourisation on Animation Videos

    cs.CV 2026-08 conditional novelty 6.0 of 10

    A training-free inference framework combining target-aware reference expansion, soft top-k palette voting, and cycle-gated temporal fusion improves segment-matching colourisation on animation videos.

  3. Meta-CoT: Enhancing Granularity and Generalization in Image Editing

    cs.CV 2026-04 unverdicted novelty 6.0 of 10

    Meta-CoT uses two-level decomposition of editing operations into meta-tasks and a CoT consistency reward to improve granularity and generalization, reporting 15.8% gains across 21 tasks.

Reference graph

Works this paper leans on

41 extracted references · 19 linked inside Pith · cited by 3 Pith papers

  1. [1]

    Stable video diffusion: Scaling latent video diffusion models to large datasets

    Andreas Blattmann, Tim Dockhorn, Sumith Kulal, Daniel Mendelevitch, Maciej Kilian, Dominik Lorenz, Yam Levi, Zion English, Vikram V oleti, Adam Letts, et al. Stable video diffusion: Scaling latent video diffusion models to large datasets. arXiv preprint arXiv:2311.15127, 2023. 3

    arXiv 2023

  2. [2]

    Align your latents: High-resolution video synthesis with la- tent diffusion models

    Andreas Blattmann, Robin Rombach, Huan Ling, Tim Dock- horn, Seung Wook Kim, Sanja Fidler, and Karsten Kreis. Align your latents: High-resolution video synthesis with la- tent diffusion models. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 22563–22575, 2023. 3

    2023

  3. [3]

    Image quality assessment: Unifying structure and texture similarity

    Keyan Ding, Kede Ma, Shiqi Wang, and Eero P Simoncelli. Image quality assessment: Unifying structure and texture similarity. IEEE transactions on pattern analysis and ma- chine intelligence, 44(5):2567–2581, 2020. 6

    2020

  4. [4]

    Scaling recti- fied flow transformers for high-resolution image synthesis

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas M ¨uller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, et al. Scaling recti- fied flow transformers for high-resolution image synthesis. In Forty-first international conference on machine learning,

  5. [5]

    Cameractrl: Enabling camera control for text-to-video generation

    Hao He, Yinghao Xu, Yuwei Guo, Gordon Wetzstein, Bo Dai, Hongsheng Li, and Ceyuan Yang. Cameractrl: Enabling camera control for text-to-video generation. arXiv preprint arXiv:2404.02101, 2024. 3

    arXiv 2024

  6. [6]

    Denoising diffu- sion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models. arXiv preprint arXiv:2006.11239,

    arXiv 2006

  7. [7]

    Video dif- fusion models

    Jonathan Ho, Tim Salimans, Alexey Gritsenko, William Chan, Mohammad Norouzi, and David J Fleet. Video dif- fusion models. Advances in Neural Information Processing Systems, 35:8633–8646, 2022. 3

    2022

  8. [8]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, et al. Lora: Low-rank adaptation of large language models. ICLR, 1(2):3, 2022. 5, 8, 10

    2022

Show all 41 references
  1. [9]

    Animate anyone: Consistent and controllable image- to-video synthesis for character animation

    Li Hu. Animate anyone: Consistent and controllable image- to-video synthesis for character animation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 8153–8163, 2024. 3

    2024

  2. [10]

    In-context lora for diffusion transformers

    Lianghua Huang, Wei Wang, Zhi-Fan Wu, Yupeng Shi, Huanzhang Dou, Chen Liang, Yutong Feng, Yu Liu, and Jin- gren Zhou. In-context lora for diffusion transformers. arXiv preprint arXiv:2410.23775, 2024. 6

    2024 arXiv

  3. [11]

    Vbench: Comprehensive bench- mark suite for video generative models

    Ziqi Huang, Yinan He, Jiashuo Yu, Fan Zhang, Chenyang Si, Yuming Jiang, Yuanhan Zhang, Tianxing Wu, Qingyang Jin, Nattapol Chanpaisit, et al. Vbench: Comprehensive bench- mark suite for video generative models. In Proceedings of the IEEE/CVF Conference on Computer Vision and P...

    2024

  4. [12]

    Lvcd: reference-based lineart video colorization with diffusion models

    Zhitong Huang, Mohan Zhang, and Jing Liao. Lvcd: reference-based lineart video colorization with diffusion models. ACM Transactions on Graphics (TOG), 43(6):1–11,

  5. [13]

    Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A. Efros. Image-to-image translation with conditional adver- sarial networks. arXiv preprint arXiv:1611.07004, 2017. 3

    2017 arXiv

  6. [14]

    Exploring the frontiers of animation video gener- ation in the sora era: Method, dataset and benchmark

    Yudong Jiang, Baohan Xu, Siqian Yang, Mingyu Yin, Jing Liu, Chao Xu, Siqi Wang, Yidi Wu, Bingwen Zhu, Jixuan Xu, et al. Exploring the frontiers of animation video gener- ation in the sora era: Method, dataset and benchmark. arXiv preprint arXiv:2412.10255, 2024. 2, 3

    2024 arXiv

  7. [15]

    Hunyuanvideo: A systematic framework for large video generative models

    Weijie Kong, Qi Tian, Zijian Zhang, Rox Min, Zuozhuo Dai, Jin Zhou, Jiangfeng Xiong, Xin Li, Bo Wu, Jianwei Zhang, et al. Hunyuanvideo: A systematic framework for large video generative models. arXiv preprint arXiv:2412.03603, 2024. 3, 5

    2024 arXiv

  8. [16]

    Xiaoyu Li, Bo Zhang, Jing Liao, and Pedro V . Sander. Deep sketch-guided cartoon video inbetweening. IEEE Transac- tions on Visualization and Computer Graphics, 28(8):2938– 2952, 2021. 3

    2021

  9. [17]

    Image conductor: Precision control for interactive video syn- thesis

    Yaowei Li, Xintao Wang, Zhaoyang Zhang, Zhouxia Wang, Ziyang Yuan, Liangbin Xie, Yuexian Zou, and Ying Shan. Image conductor: Precision control for interactive video syn- thesis. arXiv preprint arXiv:2406.15339, 2024. 3

    2024 arXiv

  10. [18]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. 7

    2017 arXiv

  11. [19]

    Anidoc: Animation creation made easier

    Yihao Meng, Hao Ouyang, Hanlin Wang, Qiuyu Wang, Wen Wang, Ka Leong Cheng, Zhiheng Liu, Yujun Shen, and Huamin Qu. Anidoc: Animation creation made easier. arXiv preprint arXiv:2412.14173, 2024. 2, 3, 7

    2024 arXiv

  12. [20]

    Scalable diffusion models with transformers

    William Peebles and Saining Xie. Scalable diffusion models with transformers. arXiv preprint arXiv:2303.12345, 2023. 3

    2023 arXiv

  13. [21]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  14. [22]

    Zero: Memory optimizations toward training trillion parameter models

    Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He. Zero: Memory optimizations toward training trillion parameter models. In SC20: International Confer- ence for High Performance Computing, Networking, Storage and Analysis, pages 1–16. IEEE, 2020. 7

    2020

  15. [23]

    Enhanced deep animation video interpolation

    Wang Shen, Cheng Ming, Wenbo Bao, Guangtao Zhai, Li Chenn, and Zhiyong Gao. Enhanced deep animation video interpolation. In 2022 IEEE International Conference on Im- age Processing (ICIP), pages 31–35. IEEE, 2022. 3

    2022

  16. [24]

    Tiny and efficient model for the edge detec- tion generalization

    Xavier Soria, Yachuan Li, Mohammad Rouhani, and An- gel D Sappa. Tiny and efficient model for the edge detec- tion generalization. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 1364–1373,

  17. [25]

    Roformer: Enhanced transformer with rotary position embedding

    Jianlin Su, Murtadha Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. Roformer: Enhanced transformer with rotary position embedding. Neurocomputing, 568:127063,

  18. [26]

    Generative ai for cel-animation: A survey

    Yunlong Tang, Junjia Guo, Pinxin Liu, Zhiyuan Wang, Hang Hua, Jia-Xing Zhong, Yunzhong Xiao, Chao Huang, Luchuan Song, Susan Liang, Yizhi Song, Liu He, Jing Bi, Mingqian Feng, Xinyang Li, Zeliang Zhang, and Chenliang Xu. Generative ai for cel-animation: A survey. arXiv preprin...

    2025

  19. [27]

    Wan: Open and advanced large-scale video gen- erative models

    Ang Wang, Baole Ai, Bin Wen, Chaojie Mao, Chen-Wei Xie, Di Chen, Feiwu Yu, Haiming Zhao, Jianxiao Yang, Jianyuan Zeng, et al. Wan: Open and advanced large-scale video gen- erative models. arXiv preprint arXiv:2503.20314, 2025. 2, 3, 4, 5

    2025 arXiv

  20. [28]

    Cinemaster: A 3d-aware and controllable frame- work for cinematic text-to-video generation

    Qinghe Wang, Yawen Luo, Xiaoyu Shi, Xu Jia, Huchuan Lu, Tianfan Xue, Xintao Wang, Pengfei Wan, Di Zhang, and Kun Gai. Cinemaster: A 3d-aware and controllable frame- work for cinematic text-to-video generation. arXiv preprint arXiv:2502.08639, 2025. 3

    2025 arXiv

  21. [29]

    Cogvlm: Visual expert for pretrained language models

    Weihan Wang, Qingsong Lv, Wenmeng Yu, Wenyi Hong, Ji Qi, Yan Wang, Junhui Ji, Zhuoyi Yang, Lei Zhao, Song XiX- uan, et al. Cogvlm: Visual expert for pretrained language models. Advances in Neural Information Processing Sys- tems, 37:121475–121499, 2024. 6

    2024

  22. [30]

    Motionctrl: A unified and flexible motion controller for video generation

    Zhouxia Wang, Ziyang Yuan, Xintao Wang, Yaowei Li, Tianshui Chen, Menghan Xia, Ping Luo, and Ying Shan. Motionctrl: A unified and flexible motion controller for video generation. In ACM SIGGRAPH 2024 Conference Pa- pers, pages 1–11, 2024. 3

    2024

  23. [31]

    Adversarial open domain adap- tation for sketch-to-photo synthesis

    Xiaoyu Xiang, Ding Liu, Xiao Yang, Yiheng Zhu, Xiaohui Shen, and Jan P Allebach. Adversarial open domain adap- tation for sketch-to-photo synthesis. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, pages 1434–1444, 2022. 6, 7

    2022

  24. [32]

    Toon- crafter: Generative cartoon interpolation

    Jinbo Xing, Hanyuan Liu, Menghan Xia, Yong Zhang, Xintao Wang, Ying Shan, and Tien-Tsin Wong. Toon- crafter: Generative cartoon interpolation. arXiv preprint arXiv:2405.12345, 2024. 2, 3, 4, 5, 7

    2024 arXiv

  25. [33]

    Make-your-video: Cus- tomized video generation using textual and structural guid- ance

    Jinbo Xing, Menghan Xia, Yuxin Liu, Yuechen Zhang, Yong Zhang, Yingqing He, Hanyuan Liu, Haoxin Chen, Xi- aodong Cun, Xintao Wang, et al. Make-your-video: Cus- tomized video generation using textual and structural guid- ance. IEEE Transactions on Visualization and Computer Gra...

    2024

  26. [34]

    Dynamicrafter: Animating open-domain images with video diffusion priors

    Jinbo Xing, Menghan Xia, Yong Zhang, Haoxin Chen, Wangbo Yu, Hanyuan Liu, Gongye Liu, Xintao Wang, Ying Shan, and Tien-Tsin Wong. Dynamicrafter: Animating open-domain images with video diffusion priors. In Eu- ropean Conference on Computer Vision , pages 399–417. Springer, 2024. 3

    2024

  27. [35]

    Cogvideox: Text-to-video diffusion models with an expert transformer

    Zhuoyi Yang, Jiayan Teng, Wendi Zheng, Ming Ding, Shiyu Huang, Jiazheng Xu, Yuanming Yang, Wenyi Hong, Xiao- han Zhang, Guanyu Feng, et al. Cogvideox: Text-to-video diffusion models with an expert transformer. arXiv preprint arXiv:2408.06072, 2024. 3, 5

    2024 arXiv

  28. [36]

    Ip- adapter: Text compatible image prompt adapter for text-to- image diffusion models

    Hu Ye, Jun Zhang, Sibo Liu, Xiao Han, and Wei Yang. Ip- adapter: Text compatible image prompt adapter for text-to- image diffusion models. arXiv preprint arXiv:2308.06721,

    arXiv

  29. [37]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. arXiv preprint arXiv:2302.05543, 2023. 3, 6

    2023 arXiv

  30. [38]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 586–595, 2018. 6

    2018

  31. [39]

    Champ: Controllable and consistent human image an- imation with 3d parametric guidance

    Shenhao Zhu, Junming Leo Chen, Zuozhuo Dai, Zilong Dong, Yinghui Xu, Xun Cao, Yao Yao, Hao Zhu, and Siyu Zhu. Champ: Controllable and consistent human image an- imation with 3d parametric guidance. In European Confer- ence on Computer Vision , pages 145–162. Springer, 2024. 3

    2024

  32. [40]

    Colorflow: Retrieval- augmented image sequence colorization

    Junhao Zhuang, Xuan Ju, Zhaoyang Zhang, Yong Liu, Shiyi Zhang, Chun Yuan, and Ying Shan. Colorflow: Retrieval- augmented image sequence colorization. arXiv preprint arXiv:2412.11815, 2024. 3

    2024 arXiv

  33. [41]

    Cobra: Efficient line art colorization with broader references

    Junhao Zhuang, Lingen Li, Xuan Ju, Zhaoyang Zhang, Chun Yuan, and Ying Shan. Cobra: Efficient line art colorization with broader references. arXiv preprint arXiv:2504.12240,

    arXiv

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.