Pith. sign in

REVIEW 4 major objections 202 references

Motion4Motion: Motion Transfer Across Subjects at Inference

T0 review · 4 major / 0 minor · reviewed 2026-07-14 · grok-4.5

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

Pith's one-line read Motion can be transferred across subjects of any shape at inference by injecting pixel motion flows into a video diffusion model, without skeletons or training.

desk verdict Clean training-free TransPE recipe that actually moves the needle on cross-species video motion transfer; the matching step is the real soft underbelly. read the letter →

arxiv 2607.11644 v1 pith:6UFDSBR2 submitted 2026-07-13 cs.CV

Ling-Hao Chen , Zixin Yin , Duomin Wang , Xianfang Zeng , Gang Yu This is my paper
classification cs.CV
keywords motiontransfervideodiffusionattentioncontroltraining-freecross-speciesanimationflowTransPETransformer
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

The paper claims that motion transfer between videos does not need human skeletons or large-scale retraining. Extract dense motion flows from a source video, match a few semantic points to a target subject, and inject those trajectories into the self-attention of a pretrained diffusion transformer so the target inherits the source motion. The training-free system, Motion4Motion, works for animals of different species and even for inanimate objects such as a walking table. Animation has long been locked to fixed body structures; removing that lock would let creators drive arbitrary characters from ordinary reference videos while keeping identity and texture intact.

What carries the argument

TransPE (Transferring Positional Encoding): during denoising the query stays unchanged while key and value tensors are concatenated with the target subject’s appearance features, re-embedded by rotary positional encoding along the source motion trajectories; this places the target’s look at the coordinates dictated by the source flow.

What would settle it

On a held-out set of cross-species pairs with ground-truth part labels (e.g., dog legs to giraffe legs), if matched points systematically land on wrong parts and the generated videos show limbs moving out of semantic correspondence while skeleton-based methods do not, the claim that skeleton-free flow injection is sufficient would fail.

Watch

Extended reading notes

Core claim

Motion4Motion is a training-free framework that achieves high-fidelity motion transfer across general subjects, including different species and non-human morphologies, by representing motion as dense pixel-level flows rather than skeletons and injecting those flows into a diffusion transformer’s self-attention via Transferring Positional Encoding (TransPE), with no model fine-tuning required.

Load-bearing premise

The method assumes automatic point matching and tracking reliably link corresponding body parts across very different shapes, so the injected trajectories land on the right places of the target.

Editorial extensions

If this is right

  • Creators can animate non-human or non-animal subjects from ordinary reference videos without building new skeletons.
  • Cross-morphology concept composition (for example a desk that walks) becomes practical by binding semantic parts with masks.
  • Existing video diffusion models can be used for motion control without collecting topology-specific paired data.
  • Per-video optimization and skeleton detectors are unnecessary for many retargeting tasks.

Reading between the lines

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

  • If point matching remains the bottleneck, combining sparse learned correspondences with TransPE could bridge larger morphological gaps.
  • The same attention-padding pattern could move other attributes (style, lighting) by choosing which features are re-positioned.
  • Systematic failure cases on heavy occlusion or extreme articulation would map the practical limits of training-free diffusion control more clearly than current animal/human benchmarks.
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

4 major / 0 minor

Summary. The paper proposes Motion4Motion, a training-free inference-time method for transferring motion from a source video to a target subject (including cross-species and non-human morphologies) without skeletal priors. It extracts dense motion flows via SAM-2 anchors, DIFT-based cross-image matching, and point tracking, then injects the resulting target trajectories into a WAN DiT’s self-attention through a TransPE module that concatenates RoPE-repositioned appearance keys/values (Eqs. 2–3, Alg. 1). Experiments on human/animal splits, a new 50-pair pose-similarity set, a 10-rater user study, and a table-walking application report gains over FlexiAct, MotionClone, RoPECraft, and MotionDirector on TS/MF/TC/AC/PS.

Significance. Removing the shared-skeleton assumption for video motion transfer is a practically important step for animation and digital creation with diverse characters. A training-free DiT attention intervention that can retarget motion across species and, with care, to inanimate objects would be a useful tool for the community. Strengths include a clear algorithmic pipeline (Alg. 1), quantitative tables plus a blind pairwise user study, and an explicit novel-composition application. The contribution is primarily engineering and empirical rather than theoretical; its lasting value hinges on whether the correspondence stage is reliable enough to support the “general subjects” claim.

major comments (4)
  1. Sec. 3.2–3.3 and the composite mapping M = I ∘ C⁻¹: the central claim of morphology-agnostic transfer rests on DIFT semantic matching of SAM-2 anchors plus DIFT tracking. The manuscript never reports matching precision, failure rate, or correspondence error as a function of morphological distance (dog↔lion, fox↔giraffe, human legs↔table). Without this, the large PS gain in Table 1 (0.543) and the “general subjects” claim cannot be attributed cleanly to TransPE rather than to favorable matches on the chosen pairs.
  2. Sec. 5 (table-walking application): the paper explicitly resorts to manual SAM-2 “bone binding” masks to make the human→table case tractable, stating that pure matching is insufficient under large structural disparity. This is a load-bearing admission: the flagship cross-morphology result is not fully automatic. The abstract and contribution bullets should state the degree of manual correspondence required, and automatic vs. bound results should be separated in evaluation.
  3. Sec. 4.1.2 vs. Table 1: the text claims comparison with training-required trajectory methods Diffusion-As-Shader and WAN-Move, yet Table 1 reports only MotionDirector, RoPECraft, MotionClone, and FlexiAct. Either add those numbers under a fair protocol (or explain why they are inapplicable) or remove the claim of outperforming them; otherwise the “outperform baselines impressively” statement is incomplete.
  4. Sec. 4.1.1 free parameters (manipulation step range [0,35], layer range [0,40], number of anchors N): Fig. 7 ablates only the step cutoff. There is no sensitivity analysis for N or layer range, nor any failure-mode analysis when matching is noisy. Because these knobs directly control where appearance is injected (Eqs. 2–3), a short ablation or default-robustness check is needed to support reproducibility of the Table 1 gains.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: training-free attention construction evaluated on external benchmarks, not a self-referential derivation.

full rationale

Motion4Motion presents an engineering pipeline (DIFT/SAM-2 correspondence + motion-flow extraction + TransPE key/value padding inside WAN DiT self-attention, Eqs. 2–3 and Alg. 1) that is applied at inference and scored against external video pairs, Co-tracker trajectories, CLIP metrics, a pose detector, and human raters (Table 1, Sec. 4). No parameter is fitted to a subset of the evaluation data and then re-presented as a prediction; no uniqueness theorem or ansatz is imported from the authors’ prior work to force the numerical scores; the composite mapping M = I ∘ C⁻¹ is an operational definition of the control signal, not a claimed first-principles derivation that reduces to its own inputs. Self-citations (e.g., Motion2Motion, HumanMAC) appear only as background related work and do not underwrite the central empirical claims. The method is therefore self-contained against external benchmarks and exhibits zero circularity of the kinds enumerated.

Assumptions & free parameters 3 free parameters · 3 assumptions · 1 invented entities

As an empirical systems paper the load-bearing items are implementation choices and modeling assumptions rather than free physical constants. The free parameters are the hand-chosen ranges for attention intervention; the axioms are the standard diffusion and correspondence assumptions; TransPE is the sole invented module.

free parameters (3)
  • attention-manipulation step range = steps 0–35
    Default [0,35] out of 50 denoising steps; chosen by ablation (Fig. 7) and directly affects motion-vs-appearance trade-off.
  • attention-manipulation layer range = layers 0–40
    Default layers [0,40]; hand-selected and not derived from first principles.
  • number of anchor points N
    Number of sampled points for matching/tracking; affects density of motion control but is not fixed by theory.
assumptions (3)
  • domain assumption 3D RoPE makes self-attention locally position-sensitive, so re-encoding keys with new trajectories relocates appearance features.
    Invoked in Sec. 3.1–3.3 and Fig. 2; standard for RoPE-based DiTs but not proven for arbitrary morphology gaps.
  • domain assumption Diffusion features (DIFT) plus SAM-2 yield sufficiently accurate semantic correspondences across species.
    Core of Sec. 3.2; if false the composite mapping M fails.
  • domain assumption Deterministic inversion of a generated or I2V video recovers a usable noise latent for subsequent editing.
    Sec. 3.2 and 4.1; relies on UniEdit-Flow-style inversion quality.
invented entities (1)
  • TransPE (Transferring Positional Encoding)
    purpose: Pads key/value tensors with target appearance features whose RoPE coordinates follow the source motion flow, thereby relocating the subject without training.
    Defined in Sec. 3.3 and Eq. (2)–(3); the sole novel algorithmic primitive. Independent evidence is only the empirical tables and figures of this paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Motion4Motion: Motion Transfer Across Subjects at Inference." pith.science (2026). https://pith.science/paper/6UFDSBR2

@misc{pith2026260711644,
  author       = {Pith},
  title        = {Pith review of: Motion4Motion: Motion Transfer Across Subjects at Inference},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6UFDSBR2}},
  note         = {Machine review of arXiv:2607.11644}
}
read the original abstract

This work explores the motion transfer from one video to another, which is crucial in animation for diverse characters. Previously, video motion transfer has been largely explored between human and human-like characters, enabling a lot of applications in digital creation. However, these approaches encounter a main limitation. Specifically, related technical pipelines heavily rely on a predefined human skeleton structure and accordingly require skeleton-conditional model training. On the one hand, these methods are difficult to generalize to diverse characters, such as animals from different species, while preserving their unique motion styles. On the other hand, labeled data in diverse skeletons is limited, which additionally restricts the large-scale training for the task. In this paper, we jump out of the skeleton-based motion transfer framework and propose a training-free motion transfer framework, named Motion4Motion. Motion4Motionmodels the motion flow of the character in a video instead of skeletons, which makes motion transfer across species easier. Extensive experimental results and novel applications show our methods outperform baselines impressively. Project page is available at https://lhchen.top/Motion4Motion.

Figures

Figures reproduced from arXiv: 2607.11644 by the authors.

Figure 1
Figure 1. Introducing Motion4Motion, a framework transferring motion from one subject to another. Motion4Motion achieves cross-species (e.g., human → panda, or human → goose) motion transfer without a uniform skeleton at inference. This work explores the motion transfer from one video to another, which is crucial in animation for diverse characters. Previously, video motion trans￾fer has been largely explored between human an… view at source ↗
Figure 2
Figure 2. Position awareness in self-attention. We sample two frames from a video. The source point in (A) is marked by a point ★. We found the attention weight (overlaid on (C)) of this point ★ is more aware of the spatial-temporal neighbors ★, but not semantically similar ones ▲. Matching [Lipman et al. 2023]. For efficient processing, an input video is first compressed into a latent space z ∈ R (1+⌊ 𝐹 4 ⌋ )× 𝐻 8 × 𝑊 8 ×𝐶 v… view at source ↗
Figure 3
Figure 3. Building correspondences across images and video. (A) Cross￾image correspondence: Anchor points P 1 𝑠𝑟𝑐 sampled within the source mask are semantically matched to the target subject P𝑡𝑔𝑡 via point matching. (B) Motion flow extraction: The motion flow M𝑠𝑟𝑐 is constructed by tracking the trajectories of these points across the temporal axis of the source video. encoding plays in video generation. Such a relative dista… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: System overview of Motion4Motion. (A) Overall Framework. Our system adds standard self-attention with the TransPE module within DiT blocks to achieve training-free motion transfer. (B) Mechanism of TransPE. The original Query Q (•) keeps unchanged. Key K (♦) is padded …
Figure 5
Figure 5. Figure 5: Comparison of Motion4Motion with baselines. We focus on the preservation of pose details across individual frames. Most baseline methods exhibit varying degrees of visual artifacts or appearance drifting. In contrast, our method maintains high-fidelity pose alignment a…
Figure 7
Figure 7. Figure 7: Ablation on manipulation steps. (A) Without manipulation (Step 0), the model fails to relocate target features along the intended trajectories. (B) Our default strategy (Step 35) achieves an optimal balance between motion fidelity and visual realism. (C) Full-step mani…
Figure 8
Figure 8. Figure 8: Application of Motion4Motion in Cross-domain Motion Synthesis. We empower a vanilla T2V model with the capability to animate inanimate objects using biological motion trajectories. Given a reference video of a walking person (A) and a target static image of a table (B)…
Figure 9
Figure 9. Figure 9: Examples of the motion transfer results bt Motion4Motion. SIGGRAPH Conference Papers ’26, July 19–23, 2026, Los Angeles, CA, USA [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

202 extracted references · 13 linked inside Pith

  1. [1]

    Abril and Robert Plant

    Patricia S. Abril and Robert Plant. The patent holder's dilemma: Buy, sell, or troll?. Communications of the ACM. doi:10.1145/1188913.1188915

    doi:10.1145/1188913.1188915

  2. [2]

    Deciding equivalances among conjunctive aggregate queries

    Sarah Cohen and Werner Nutt and Yehoshua Sagic. Deciding equivalances among conjunctive aggregate queries. doi:10.1145/1219092.1219093

    doi:10.1145/1219092.1219093

  3. [3]

    Special issue: Digital Libraries. 1996

    1996

  4. [4]

    Understanding Policy-Based Networking

    David Kosiur. Understanding Policy-Based Networking

  5. [7]

    doi:10.1007/3-540-09237-4

    The title of book two. doi:10.1007/3-540-09237-4

    doi:10.1007/3-540-09237-4

  6. [8]

    Asad Z. Spector. Achieving application requirements. Distributed Systems. doi:10.1145/90417.90738

    doi:10.1145/90417.90738

  7. [9]

    Douglass and David Harel and Mark B

    Bruce P. Douglass and David Harel and Mark B. Trakhtenbrot. Statecarts in use: structured analysis and object-orientation. Lectures on Embedded Systems. doi:10.1007/3-540-65193-4_29

    doi:10.1007/3-540-65193-4_29

  8. [10]

    Donald E. Knuth. The Art of Computer Programming, Vol. 1: Fundamental Algorithms (3rd. ed.)

Show all 202 references
  1. [11]

    Donald E. Knuth. The Art of Computer Programming

  2. [12]

    Structured Variational Inference Procedures and their Realizations (as incol)

    Dan Geiger and Christopher Meek. Structured Variational Inference Procedures and their Realizations (as incol). Proceedings of Tenth International Workshop on Artificial Intelligence and Statistics, The Barbados

  3. [13]

    Stan W. Smith. An experiment in bibliographic mark-up: Parsing metadata for XML export. Proceedings of the 3rd. annual workshop on Librarians and Computers

  4. [14]

    Catch me, if you can: Evading network signatures with web-based polymorphic worms

    Matthew Van Gundy and Davide Balzarotti and Giovanni Vigna. Catch me, if you can: Evading network signatures with web-based polymorphic worms. Proceedings of the first USENIX workshop on Offensive Technologies

  5. [15]

    Predicate Path expressions

    Sten Andler. Predicate Path expressions. Proceedings of the 6th. ACM SIGACT-SIGPLAN symposium on Principles of Programming Languages. doi:10.1145/567752.567774

    doi

  6. [16]

    LOGICS of Programs: AXIOMATICS and DESCRIPTIVE POWER

    David Harel. LOGICS of Programs: AXIOMATICS and DESCRIPTIVE POWER

  7. [17]

    Anisi , title =

    David A. Anisi , title =

  8. [18]

    Clarkson

    Kenneth L. Clarkson. Algorithms for Closest-Point Problems (Computational Geometry)

  9. [19]

    Introduction to Bayesian Statistics

    Harry Thornburg. Introduction to Bayesian Statistics. 2001

    2001

  10. [20]

    CLIFFORD: a Maple 11 Package for Clifford Algebra Computations, version 11

    Rafal Ablamowicz and Bertfried Fauser. CLIFFORD: a Maple 11 Package for Clifford Algebra Computations, version 11. 2007

    2007

  11. [21]

    Stats and Analysis

    Poker-Edge.Com. Stats and Analysis. 2006

    2006

  12. [22]

    A more perfect union

    Barack Obama. A more perfect union

  13. [23]

    The fountain of youth

    Joseph Scientist. The fountain of youth

  14. [24]

    Solder man

    Dave Novak. Solder man. ACM SIGGRAPH 2003 Video Review on Animation theater Program: Part I - Vol. 145 (July 27--27, 2003). doi:10.945/woot07-S422

    2003

  15. [25]

    Interview with Bill Kinder: January 13, 2005

    Newton Lee. Interview with Bill Kinder: January 13, 2005. Comput. Entertain. doi:10.1145/1057270.1057278

    2005 doi

  16. [26]

    The Enabling of Digital Libraries

    Bernard Rous. The Enabling of Digital Libraries. Digital Libraries

  17. [28]

    (new) Finding minimum congestion spanning trees , journal =

    Werneck, Renato and Setubal, Jo\. (new) Finding minimum congestion spanning trees , journal =. doi:10.1145/351827.384253 , acmid = 384253, publisher =

    doi

  18. [30]

    and Mei, Alessandro , title =

    Conti, Mauro and Di Pietro, Roberto and Mancini, Luigi V. and Mei, Alessandro , title =. Inf. Fusion , volume =. 2009 , issn =. doi:10.1016/j.inffus.2009.01.002 , acmid =

    2009 doi

  19. [31]

    and Hutchful, David K

    Li, Cheng-Lun and Buyuktur, Ayse G. and Hutchful, David K. and Sant, Natasha B. and Nainwal, Satyendra K. , title =. CHI '08 extended abstracts on Human factors in computing systems , year =. doi:10.1145/1358628.1358946 , acmid =

    doi

  20. [32]

    , title =

    Hollis, Billy S. , title =. 1999 , isbn =

    1999

  21. [33]

    Goossens, Michel and Rahtz, S. P. and Moore, Ross and Sutor, Robert S. , title =. 1999 , isbn =

    1999

  22. [34]

    and Rosenberg, Arnold L

    Buss, Jonathan F. and Rosenberg, Arnold L. and Knott, Judson D. , title =. 1987 , source =

    1987

  23. [35]

    CHI '08: CHI '08 extended abstracts on Human factors in computing systems , year =

    , note =. CHI '08: CHI '08 extended abstracts on Human factors in computing systems , year =

  24. [36]

    Algorithms for Closest-Point Problems (Computational Geometry) , year =

    Clarkson, Kenneth Lee , advisor =. Algorithms for Closest-Point Problems (Computational Geometry) , year =

  25. [37]

    SIGCOMM Comput. Commun. Rev. , year =

  26. [38]

    2004 , isbn =

    IEEE TCSC Executive Committee , booktitle =. 2004 , isbn =. doi:http://dx.doi.org/10.1109/ICWS.2004.64 , acmid =

    2004 doi

  27. [39]

    Distributed systems (2nd Ed.) , year =

  28. [40]

    , title =

    Petrie, Charles J. , title =. 1986 , source =

    1986

  29. [41]

    Donald E. Knuth. Seminumerical Algorithms. 1981

    1981

  30. [42]

    E-commerce and cultural values , year =

    Kong, Wei-Chang , Title =. E-commerce and cultural values , year =

  31. [43]

    E-commerce and cultural values , year =

    Kong, Wei-Chang , type =. E-commerce and cultural values , year =

  32. [44]

    Chapter 9 , booktitle =

    Kong, Wei-Chang , editor =. Chapter 9 , booktitle =

  33. [45]

    E-commerce and cultural values , editor =

    Kong, Wei-Chang , title =. E-commerce and cultural values , editor =. 2003 , isbn =

    2003

  34. [46]

    E-commerce and cultural values - (InBook-num-in-chap) , chapter =

    Kong, Wei-Chang , editor =. E-commerce and cultural values - (InBook-num-in-chap) , chapter =. 2004 , address =

    2004

  35. [47]

    E-commerce and cultural values (Inbook-text-in-chap) , chapter =

    Kong, Wei-Chang , editor =. E-commerce and cultural values (Inbook-text-in-chap) , chapter =. 2005 , address =

    2005

  36. [48]

    E-commerce and cultural values (Inbook-num chap) , chapter =

    Kong, Wei-Chang , editor =. E-commerce and cultural values (Inbook-num chap) , chapter =. 2006 , address =

    2006

  37. [49]

    Microelectron

    Mehdi Saeedi and Morteza Saheb Zamani and Mehdi Sedighi , title =. Microelectron. J. , volume =. 2010 , pages =

    2010

  38. [50]

    Mehdi Saeedi and Morteza Saheb Zamani and Mehdi Sedighi and Zahra Sasanian , title =. J. Emerg. Technol. Comput. Syst. , volume =

  39. [51]

    Kirschmer, Markus and Voight, John , title =. SIAM J. Comput. , issue_date =. 2010 , issn =. doi:https://doi.org/10.1137/080734467 , acmid =

    2010 doi

  40. [52]

    Hoare, C. A. R. , title =. Structured programming (incoll) , editor =. 1972 , isbn =

    1972

  41. [53]

    History of programming languages I (incoll) , editor =

    Lee, Jan , title =. History of programming languages I (incoll) , editor =. 1981 , isbn =. doi:http://doi.acm.org/10.1145/800025.1198348 , acmid =

    1981 doi

  42. [54]

    , title =

    Dijkstra, E. , title =. Classics in software engineering (incoll) , year =

  43. [55]

    , title =

    Wenzel, Elizabeth M. , title =. Multimedia interface design (incoll) , year =. doi:10.1145/146022.146089 , acmid =

    doi

  44. [56]

    , title =

    Mumford, E. , title =. Critical issues in information systems research (incoll) , year =

  45. [57]

    and Golden, Donald G

    McCracken, Daniel D. and Golden, Donald G. , title =. 1990 , isbn =

    1990

  46. [58]

    The analysis of linear partial differential operators

    H. The analysis of linear partial differential operators. 1985 , PAGES =

    1985

  47. [59]

    IEEE", address =

    A. Adya and P. Bahl and J. Padhye and A.Wolman and L. Zhou , title =. Proceedings of the IEEE 1st International Conference on Broadnets Networks (BroadNets'04) , publisher = "IEEE", address = "Los Alamitos, CA", year =

  48. [60]

    I. F. Akyildiz and W. Su and Y. Sankarasubramaniam and E. Cayirci , title =. Comm. ACM , volume = 38, number = "4", year =

  49. [61]

    I. F. Akyildiz and T. Melodia and K. R. Chowdhury , title =. Computer Netw. , volume = 51, number = "4", year =

  50. [62]

    ACM", address =

    P. Bahl and R. Chancre and J. Dungeon , title =. Proceeding of the 10th International Conference on Mobile Computing and Networking (MobiCom'04) , publisher = "ACM", address = "New York, NY", year =

  51. [63]

    8 (Special Issue on Sensor Networks)

    D. Culler and D. Estrin and M. Srivastava , title =. IEEE Comput. , volume = 37, number = "8 (Special Issue on Sensor Networks)", publisher = "IEEE", address = "Los Alamitos, CA", year =

  52. [64]

    Natarajan and M

    A. Natarajan and M. Motani and B. de Silva and K. Yap and K. C. Chua , title =. Network Architectures , editor =. 960935712

  53. [65]

    Tzamaloukas and J

    A. Tzamaloukas and J. J. Garcia-Luna-Aceves , title =

  54. [66]

    Zhou and J

    G. Zhou and J. Lu and C.-Y. Wan and M. D. Yarvis and J. A. Stankovic , title =

  55. [67]

    Mapping Powerlists onto Hypercubes

    Jacob Kornerup. Mapping Powerlists onto Hypercubes. 1994

    1994

  56. [68]

    Automatic Parallelization for Distributed-Memory Multiprocessing Systems

    Michael Gerndt. Automatic Parallelization for Distributed-Memory Multiprocessing Systems

  57. [69]

    J. E. Archer, Jr. and R. Conway and F. B. Schneider. User recovery and reversal in interactive systems. ACM Trans. Program. Lang. Syst

  58. [70]

    D. D. Dunlop and V. R. Basili. Generalizing specifications for uniformly implemented loops. ACM Trans. Program. Lang. Syst

  59. [71]

    Heering and P

    J. Heering and P. Klint. Towards monolingual programming environments. ACM Trans. Program. Lang. Syst

  60. [72]

    Donald E. Knuth. The book

  61. [73]

    Korach and D

    E. Korach and D. Rotem and N. Santoro. Distributed algorithms for finding centers and medians in networks. ACM Trans. Program. Lang. Syst

  62. [74]

    : A Document Preparation System

    Leslie Lamport. : A Document Preparation System

  63. [75]

    F. Nielson. Program transformations in a denotational setting. ACM Trans. Program. Lang. Syst

  64. [76]

    Brian K. Reid. A high-level approach to computer document formatting. Proceedings of the 7th Annual Symposium on Principles of Programming Languages

  65. [77]

    and Abdelzaher, Tarek F

    Zhou, Gang and Wu, Yafeng and Yan, Ting and He, Tian and Huang, Chengdu and Stankovic, John A. and Abdelzaher, Tarek F. , title =. ACM Trans. Embed. Comput. Syst. , issue_date =. doi:10.1145/1721695.1721705 , acmid = 1721705, publisher =

    doi

  66. [78]

    Institutional members of the Users Group

  67. [79]

    Boris Veytsman , title =

  68. [80]

    Robin Schneider , title =

  69. [81]

    and Peterson, Larry L

    Bowman, Mic and Debray, Saumya K. and Peterson, Larry L. , title =. ACM Trans. Program. Lang. Syst. , volume =. 1993 , doi =

    1993

  70. [82]

    TUGboat , volume =

    Braams, Johannes , title =. TUGboat , volume =

  71. [83]

    Post Congress Tristesse

    Malcolm Clark. Post Congress Tristesse. TeX90 Conference Proceedings

  72. [84]

    ACM Trans

    Herlihy, Maurice , title =. ACM Trans. Program. Lang. Syst. , volume =. 1993 , doi =

    1993

  73. [85]

    Salas and Einar Hille

    S.L. Salas and Einar Hille. Calculus: One and Several Variable. 1978

    1978

  74. [86]

    Publication quality tables in

    Simon Fear , month =. Publication quality tables in

  75. [87]

    Using the amsthm Package , organization =

  76. [88]

    2019 , url =

    R: A Language and Environment for Statistical Computing , author =. 2019 , url =

    2019

  77. [89]

    Sam Anzaroot and Andrew McCallum , title =

  78. [90]

    Brad and Haunschild, Robin , title =

    Bornmann, Lutz and Wray, K. Brad and Haunschild, Robin , title =

  79. [91]

    2014 , archivePrefix =

    Sam Anzaroot and Alexandre Passos and David Belanger and Andrew McCallum , title =. 2014 , archivePrefix =. 1403.1349 , doi =

    2014 arXiv

  80. [92]

    Proceedings of the 20th International Colloquium on Automata, Languages and Programming , series =

    Maintaining Discrete Probability Distributions Optimally , author =. Proceedings of the 20th International Colloquium on Automata, Languages and Programming , series =

  81. [93]

    Reiser , year = 2014, title =

    Brian J. Reiser , year = 2014, title =

    2014

  82. [94]

    Baggett and M

    R. Baggett and M. Simecek and C. Chambellan and K. Tsui and M. Fraune , year = 2025, title =

    2025

  83. [95]

    CVPR , pages=

    Rave: Randomized noise shuffling for fast and consistent video editing with diffusion models , author=. CVPR , pages=

  84. [96]

    NeurIPS , volume=

    Cove: Unleashing the diffusion feature correspondence for consistent video editing , author=. NeurIPS , volume=

  85. [97]

    ICLR , year=

    TokenFlow: Consistent Diffusion Features for Consistent Video Editing , author=. ICLR , year=

  86. [98]

    ICLR , year=

    FLATTEN: optical FLow-guided ATTENtion for consistent text-to-video editing , author=. ICLR , year=

  87. [99]

    ICCV , pages=

    Fatezero: Fusing attentions for zero-shot text-based video editing , author=. ICCV , pages=

  88. [100]

    arXiv preprint arXiv:2503.07598 , year=

    Vace: All-in-one video creation and editing , author=. arXiv preprint arXiv:2503.07598 , year=

    arXiv

  89. [101]

    CVPR , pages=

    Vbench: Comprehensive benchmark suite for video generative models , author=. CVPR , pages=

  90. [102]

    ICLR , year=

    Noise Map Guidance: Inversion with Spatial Context for Real Image Editing , author=. ICLR , year=

  91. [103]

    International Conference on Learning Representations , year=

    SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations , author=. International Conference on Learning Representations , year=

  92. [104]

    arXiv preprint arXiv:2504.13109 , year=

    UniEdit-Flow: Unleashing Inversion and Editing in the Era of Flow Models , author=. arXiv preprint arXiv:2504.13109 , year=

    arXiv

  93. [105]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Ditctrl: Exploring attention control in multi-modal diffusion transformer for tuning-free multi-prompt longer video generation , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  94. [106]

    Forty-second ICML , year=

    Taming Rectified Flow for Inversion and Editing , author=. Forty-second ICML , year=

  95. [107]

    Forty-second ICML , year=

    FireFlow: Fast Inversion of Rectified Flow for Image Semantic Editing , author=. Forty-second ICML , year=

  96. [108]

    ICLR , year=

    PnP Inversion: Boosting Diffusion-based Editing with 3 Lines of Code , author=. ICLR , year=

  97. [109]

    CVPR , pages=

    Splicing vit features for semantic appearance transfer , author=. CVPR , pages=

  98. [110]

    IEEE Transactions on pattern analysis and machine intelligence , volume=

    A computational approach to edge detection , author=. IEEE Transactions on pattern analysis and machine intelligence , volume=. 1986 , publisher=

    1986

  99. [111]

    IEEE transactions on image processing , volume=

    Image quality assessment: from error visibility to structural similarity , author=. IEEE transactions on image processing , volume=. 2004 , publisher=

    2004

  100. [112]

    ICML , pages=

    Learning transferable visual models from natural language supervision , author=. ICML , pages=. 2021 , organization=

    2021

  101. [113]

    Forty-first ICML , year=

    Scaling rectified flow transformers for high-resolution image synthesis , author=. Forty-first ICML , year=

  102. [114]

    The Eleventh International Conference on Learning Representations , year=

    DiffEdit: Diffusion-based semantic image editing with mask guidance , author=. The Eleventh International Conference on Learning Representations , year=

  103. [115]

    CoRR , year=

    Inversion-Free Image Editing with Natural Language , author=. CoRR , year=

  104. [116]

    CVPR , pages=

    An edit friendly ddpm noise space: Inversion and manipulations , author=. CVPR , pages=

  105. [117]

    ICCV , pages=

    Masactrl: Tuning-free mutual self-attention control for consistent image synthesis and editing , author=. ICCV , pages=

  106. [118]

    CVPR , pages=

    Plug-and-play diffusion features for text-driven image-to-image translation , author=. CVPR , pages=

  107. [119]

    Journal of machine learning research , volume=

    Exploring the limits of transfer learning with a unified text-to-text transformer , author=. Journal of machine learning research , volume=

  108. [120]

    2023 , howpublished =

    Damian Perry , title =. 2023 , howpublished =

    2023

  109. [121]

    2024 , note =

    Black Forest Labs , title =. 2024 , note =

    2024

  110. [122]

    2024 , note =

    Stability AI , title =. 2024 , note =

    2024

  111. [123]

    ICLR , year=

    CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer , author=. ICLR , year=

  112. [124]

    arXiv preprint arXiv:2509.06155 , year=

    UniVerse-1: Unified Audio-Video Generation via Stitching of Experts , author=. arXiv preprint arXiv:2509.06155 , year=

    arXiv

  113. [125]

    arXiv preprint arXiv:2508.09131 , year=

    Training-Free Text-Guided Color Editing with Multi-Modal Diffusion Transformer , author=. arXiv preprint arXiv:2508.09131 , year=

  114. [126]

    ICCV , pages=

    Characonsist: Fine-grained consistent character generation , author=. ICCV , pages=

  115. [127]

    Proceedings of the SIGGRAPH Asia 2025 Conference Papers , pages=

    ConsistEdit: Highly Consistent and Precise Training-free Visual Editing , author=. Proceedings of the SIGGRAPH Asia 2025 Conference Papers , pages=

    2025

  116. [128]

    Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers , pages=

    Flexiact: Towards flexible action control in heterogeneous scenarios , author=. Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers , pages=

  117. [129]

    CVPR , pages=

    Animate anyone: Consistent and controllable image-to-video synthesis for character animation , author=. CVPR , pages=

  118. [130]

    Forty-second ICML , year=

    MimicMotion: High-Quality Human Motion Video Generation with Confidence-aware Pose Guidance , author=. Forty-second ICML , year=

  119. [131]

    arXiv preprint arXiv:2509.14055 , year=

    Wan-animate: Unified character animation and replacement with holistic replication , author=. arXiv preprint arXiv:2509.14055 , year=

  120. [132]

    European Conference on Computer Vision , pages=

    Motiondirector: Motion customization of text-to-video diffusion models , author=. European Conference on Computer Vision , pages=. 2024 , organization=

    2024

  121. [133]

    ICLR , year=

    MotionClone: Training-Free Motion Cloning for Controllable Video Generation , author=. ICLR , year=

  122. [134]

    2025 , eprint=

    MotionV2V: Editing Motion in a Video , author=. 2025 , eprint=

    2025

  123. [135]

    arXiv preprint arXiv:2505.13344 , year=

    RoPECraft: Training-Free Motion Transfer with Trajectory-Guided RoPE Optimization on Diffusion Transformers , author=. arXiv preprint arXiv:2505.13344 , year=

  124. [136]

    arXiv preprint arXiv:2509.12203 , year=

    LazyDrag: Enabling Stable Drag-Based Editing on Multi-Modal Diffusion Transformers via Explicit Correspondence , author=. arXiv preprint arXiv:2509.12203 , year=

  125. [137]

    arXiv preprint arXiv:2503.20314 , year=

    Wan: Open and Advanced Large-Scale Video Generative Models , author=. arXiv preprint arXiv:2503.20314 , year=

    arXiv

  126. [138]

    CVPR , pages=

    Towards understanding cross and self-attention in stable diffusion for text-guided image editing , author=. CVPR , pages=

  127. [139]

    NeurIPS , volume=

    Uni-controlnet: All-in-one control to text-to-image diffusion models , author=. NeurIPS , volume=

  128. [140]

    CoRR , year=

    HunyuanVideo: A Systematic Framework For Large Video Generative Models , author=. CoRR , year=

  129. [141]

    CVPR , pages=

    High-resolution image synthesis with latent diffusion models , author=. CVPR , pages=

  130. [142]

    The Eleventh International Conference on Learning Representations , year=

    Prompt-to-Prompt Image Editing with Cross-Attention Control , author=. The Eleventh International Conference on Learning Representations , year=

  131. [143]

    CVPR , pages=

    Video-p2p: Video editing with cross-attention control , author=. CVPR , pages=

  132. [144]

    ICLR , year=

    FreeNoise: Tuning-Free Longer Video Diffusion via Noise Rescheduling , author=. ICLR , year=

  133. [145]

    ICLR , year=

    Semantic Image Inversion and Editing using Rectified Stochastic Differential Equations , author=. ICLR , year=

  134. [146]

    CVPR , pages=

    A style-based generator architecture for generative adversarial networks , author=. CVPR , pages=

  135. [147]

    Proceedings of the IEEE/CVF winter conference on applications of computer vision , pages=

    Training-free layout control with cross-attention guidance , author=. Proceedings of the IEEE/CVF winter conference on applications of computer vision , pages=

  136. [148]

    NeurIPS , volume=

    Dynamic prompt learning: Addressing cross-attention leakage for text-based image editing , author=. NeurIPS , volume=

  137. [149]

    arXiv preprint arXiv:2010.02502 , year=

    Denoising diffusion implicit models , author=. arXiv preprint arXiv:2010.02502 , year=

    2010 arXiv

  138. [150]

    CVPR , pages=

    Edict: Exact diffusion inversion via coupled transformations , author=. CVPR , pages=

  139. [151]

    Machine Intelligence Research , pages=

    Dpm-solver++: Fast solver for guided sampling of diffusion probabilistic models , author=. Machine Intelligence Research , pages=. 2025 , publisher=

    2025

  140. [152]

    NeurIPS , volume=

    Belm: Bidirectional explicit linear multi-step sampler for exact inversion in diffusion models , author=. NeurIPS , volume=

  141. [153]

    2005 , publisher=

    Digital image processing , author=. 2005 , publisher=

    2005

  142. [154]

    ICCV , pages=

    Talking head generation with probabilistic audio-to-visual diffusion priors , author=. ICCV , pages=

  143. [155]

    CVPR , pages=

    Progressive disentangled representation learning for fine-grained controllable talking head synthesis , author=. CVPR , pages=

  144. [156]

    CVPR , pages=

    Df-gan: A simple and effective baseline for text-to-image synthesis , author=. CVPR , pages=

  145. [157]

    ICML , pages=

    Generative adversarial text to image synthesis , author=. ICML , pages=. 2016 , organization=

    2016

  146. [158]

    NeurIPS , volume=

    Denoising diffusion probabilistic models , author=. NeurIPS , volume=

  147. [159]

    ICLR , year=

    AnimateDiff: Animate Your Personalized Text-to-Image Diffusion Models without Specific Tuning , author=. ICLR , year=

  148. [160]

    ICCV , pages=

    Scalable diffusion models with transformers , author=. ICCV , pages=

  149. [161]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Generative video propagation , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  150. [162]

    Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers , pages=

    LightLab: Controlling Light Sources in Images with Diffusion Models , author=. Proceedings of the Special Interest Group on Computer Graphics and Interactive Techniques Conference Conference Papers , pages=

  151. [163]

    ICLR , year=

    Scaling in-the-wild training for diffusion-based illumination harmonization and editing by imposing consistent light transport , author=. ICLR , year=

  152. [164]

    2025 , howpublished =

    FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space , author =. 2025 , howpublished =

    2025

  153. [165]

    2025 , howpublished =

    GPT 4o Image Generation , author =. 2025 , howpublished =

    2025

  154. [166]

    arXiv preprint arXiv:2401.14159 , year=

    Grounded sam: Assembling open-world models for diverse visual tasks , author=. arXiv preprint arXiv:2401.14159 , year=

    arXiv

  155. [167]

    NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications , year=

    Classifier-Free Diffusion Guidance , author=. NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications , year=

    2021

  156. [168]

    arXiv preprint arXiv:2504.17761 , year=

    Step1x-edit: A practical framework for general image editing , author=. arXiv preprint arXiv:2504.17761 , year=

    arXiv

  157. [169]

    CVPR , pages=

    Instructpix2pix: Learning to follow image editing instructions , author=. CVPR , pages=

  158. [170]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Unveil inversion and invariance in flow transformer for versatile image editing , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

  159. [171]

    ICCV , pages=

    Flowedit: Inversion-free text-based editing using pre-trained flow models , author=. ICCV , pages=

  160. [172]

    arXiv preprint arXiv:2503.10270 , year=

    Eedit: Rethinking the spatial and temporal redundancy for efficient image editing , author=. arXiv preprint arXiv:2503.10270 , year=

  161. [173]

    Proceedings of the 25th annual conference on Computer graphics and interactive techniques , pages=

    Retargetting motion to new characters , author=. Proceedings of the 25th annual conference on Computer graphics and interactive techniques , pages=

  162. [174]

    Motion in Games: 5th International Conference, MIG 2012, Rennes, France, November 15-17, 2012

    Automating the transfer of a generic set of behaviors onto a virtual character , author=. Motion in Games: 5th International Conference, MIG 2012, Rennes, France, November 15-17, 2012. Proceedings 5 , pages=. 2012 , organization=

    2012

  163. [175]

    Proceedings of the 26th annual conference on Computer graphics and interactive techniques , pages=

    A hierarchical approach to interactive motion editing for human-like figures , author=. Proceedings of the 26th annual conference on Computer graphics and interactive techniques , pages=

  164. [176]

    ACM TOG , volume=

    Skeleton-aware networks for deep motion retargeting , author=. ACM TOG , volume=. 2020 , publisher=

    2020

  165. [177]

    Computers & Graphics , volume=

    An iterative solution for improving the generalization ability of unsupervised skeleton motion retargeting , author=. Computers & Graphics , volume=. 2022 , publisher=

    2022

  166. [178]

    CVPR , pages=

    Semantics-aware motion retargeting with vision-language models , author=. CVPR , pages=

  167. [179]

    30th British Machine Vision Conference (BMVC 2019) , year=

    Pmnet: Learning of disentangled pose and movement for unsupervised motion retargeting , author=. 30th British Machine Vision Conference (BMVC 2019) , year=

    2019

  168. [180]

    SIGGRAPH Asia 2025 , year=

    Motion2Motion: Cross-topology Motion Transfer with Sparse Correspondence , author=. SIGGRAPH Asia 2025 , year=. doi:10.1145/3757377.3763811 , address=

    2025 doi

  169. [181]

    ICCV , pages=

    MotionShot: Adaptive Motion Transfer across Arbitrary Objects for Text-to-Video Generation , author=. ICCV , pages=

  170. [182]

    ICLR , year=

    Flow Matching for Generative Modeling , author=. ICLR , year=

  171. [183]

    Neurocomputing , volume=

    Roformer: Enhanced transformer with rotary position embedding , author=. Neurocomputing , volume=. 2024 , publisher=

    2024

  172. [184]

    arXiv preprint arXiv:2408.00714 , year=

    Sam 2: Segment anything in images and videos , author=. arXiv preprint arXiv:2408.00714 , year=

    arXiv

  173. [185]

    European conference on computer vision , pages=

    Grounding dino: Marrying dino with grounded pre-training for open-set object detection , author=. European conference on computer vision , pages=. 2024 , organization=

    2024

  174. [186]

    Thirty-seventh Conference on Neural Information Processing Systems , year=

    Emergent Correspondence from Image Diffusion , author=. Thirty-seventh Conference on Neural Information Processing Systems , year=

  175. [187]

    ICCV , pages=

    Imbalance in Balance: Online Concept Balancing in Generation Models , author=. ICCV , pages=

  176. [188]

    European Conference on Computer Vision , pages=

    X-pose: Detecting any keypoints , author=. European Conference on Computer Vision , pages=. 2024 , organization=

    2024

  177. [189]

    2025 , note =

    Artell , title =. 2025 , note =

    2025

  178. [190]

    Rokoko Motion Capture Solutions , howpublished =

  179. [191]

    2025 , eprint=

    UniEdit-Flow: Unleashing Inversion and Editing in the Era of Flow Models , author=. 2025 , eprint=

    2025

  180. [192]

    Proceedings of the IEEE/CVF international conference on computer vision , pages=

    Humanmac: Masked motion completion for human motion prediction , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=

  181. [193]

    arXiv preprint arXiv:2310.12978 , year=

    Humantomato: Text-aligned whole-body motion generation , author=. arXiv preprint arXiv:2310.12978 , year=

    arXiv

  182. [194]

    arXiv preprint arXiv:2410.18977 , year=

    Pay attention and move better: Harnessing attention for interactive motion generation and training-free editing , author=. arXiv preprint arXiv:2410.18977 , year=

    arXiv

  183. [195]

    European Conference on Computer Vision , pages=

    Motionlcm: Real-time controllable motion generation via latent consistency model , author=. European Conference on Computer Vision , pages=. 2024 , organization=

    2024

  184. [196]

    CVPR , pages=

    Cotracker3: Simpler and better point tracking by pseudo-labelling real videos , author=. CVPR , pages=

  185. [197]

    arXiv preprint arXiv:2501.03847 , year=

    Diffusion as Shader: 3D-aware Video Diffusion for Versatile Video Generation Control , author=. arXiv preprint arXiv:2501.03847 , year=

    arXiv

  186. [198]

    arXiv preprint arXiv:2512.08765 , year=

    Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance , author=. arXiv preprint arXiv:2512.08765 , year=

  187. [199]

    CVPR , year=

    Motion Prompting: Controlling Video Generation with Motion Trajectories , author=. CVPR , year=

  188. [200]

    CVPR , year=

    Go-with-the-Flow: Motion-Controllable Video Diffusion Models Using Real-Time Warped Noise , author=. CVPR , year=

  189. [201]

    arXiv preprint arXiv:2505.22944 , year=

    ATI: Any Trajectory Instruction for Controllable Video Generation , author=. arXiv preprint arXiv:2505.22944 , year=

    arXiv

  190. [202]

    ICLR , year=

    MotionStream: Real-Time Video Generation with Interactive Motion Controls , author=. ICLR , year=

  191. [203]

    CVPR , year=

    Style aligned image generation via shared attention , author=. CVPR , year=

  192. [204]

    CVPR , year=

    Z*: Zero-shot Style Transfer via Attention Rearrangement , author=. CVPR , year=

  193. [205]

    ACM Transactions on Graphics (TOG) , volume=

    Training-free consistent text-to-image generation , author=. ACM Transactions on Graphics (TOG) , volume=. 2024 , publisher=

    2024

  194. [206]

    ACM SIGGRAPH 2024 Conference Papers , year=

    Cross-image attention for zero-shot appearance transfer , author=. ACM SIGGRAPH 2024 Conference Papers , year=

    2024

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.