pith. sign in

arxiv: 2606.29148 · v1 · pith:VBIA3OXXnew · submitted 2026-06-28 · 💻 cs.CV · cs.AI· cs.GR· cs.LG

GPC: Large-Scale Generative Pretraining for Transferable Motor Control

Yi Shi , Yifeng Jiang , Chen Tessler , Xue Bin Peng This is my paper

Pith reviewed 2026-06-30 08:01 UTC · model grok-4.3

classification 💻 cs.CV cs.AIcs.GRcs.LG
keywords generative pretrainingmotor controlreinforcement learningtokenizationphysics simulationautoregressive transformermotion vocabularycharacter animation
0
0 comments X

The pith

A GPT-style transformer on learned motion tokens produces reusable controllers for physics-based characters.

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

The paper claims that large-scale generative pretraining creates general-purpose motor controllers by first learning a discrete motion vocabulary through joint optimization of finite scalar quantization and a reinforcement learning policy. An autoregressive transformer is then trained to model sequences of these tokens via next-token prediction, generating controls for a simulated character. A sympathetic reader would care because the resulting controller reproduces motions at very high rates, shows emergent recovery behaviors, and can be adapted to new tasks without retraining from scratch.

Core claim

The authors claim that jointly optimizing a motion vocabulary via Finite Scalar Quantization with a control policy through end-to-end reinforcement learning, followed by training a GPT-style autoregressive transformer on the resulting discrete codes, produces a generative controller that generates controls for a physically simulated character by performing next-token prediction, yielding robust general-purpose controllers for downstream applications.

What carries the argument

The autoregressive transformer that performs next-token prediction over the learned discrete motion vocabulary to output control sequences.

Load-bearing premise

That the discrete codes capture structure an autoregressive model can predict to generate controls that transfer beyond the original training motions.

What would settle it

Measuring whether an adapted controller achieves high success on a novel task whose required motions do not appear in the pretraining dataset.

Figures

Figures reproduced from arXiv: 2606.29148 by Chen Tessler, Xue Bin Peng, Yifeng Jiang, Yi Shi.

Figure 1
Figure 1. Figure 1: We present Generative Pretrained Controllers (GPC), which models a wide range of motor skills through next-token prediction with a learned discrete [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Keyframes of the results produced by the FSQ tracking controller on agile motions from Bones. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Skill quantization: An FSQ motion tracking controller maps refer￾ence motions to discrete latent tokens and decodes them to actions. The FSQ module is trained end-to-end with reinforcement learning using a motion￾tracking objective (top). Generative controller training: a transformer decoder models the skill distribution via causal self-attention, enabling autoregressive sampling of discrete skill tokens c… view at source ↗
Figure 4
Figure 4. Figure 4: Unconditional sampling of the generative controller. GPC produces a wide array of highly dynamic skills, such as jumping, leaping, and rolling. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: GPC produces robust and natural recovery behaviors when subjected to external perturbations. [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: GPC exhibits diverse and versatile responses to force perturbations with a generative controller trained with Bones. [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: CoLA can be used to efficient finetune GPC to new tasks. By only training adaptation layers during finetuning, the pretrained generative controller can [PITH_FULL_IMAGE:figures/full_fig_p006_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison between downstream task behavior of GPC and CVAE under identical task conditions. GPC produces more diverse behaviors than CVAE. [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 13
Figure 13. Figure 13: The simulated character used in experiments conducted on Bones. [PITH_FULL_IMAGE:figures/full_fig_p012_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Composited skill achieved by framewise code blending. [PITH_FULL_IMAGE:figures/full_fig_p014_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: t-SNE visualization of FSQ embeddings at various network depths. Note that the pre-quantization FSQ codes and the output embeddings of the last [PITH_FULL_IMAGE:figures/full_fig_p015_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Scenes of HSI tasks. receives a target heading direction ˆd ∈ R 2 , target speed 𝑣 ∗ , and target facing direction ˆ f ∈ R 2 . The reward combines directional velocity matching with heading alignment: 𝑟steer = 𝑤dir · 𝑟dir + 𝑤face · 𝑟face (11) where 𝑤dir = 0.7 and 𝑤face = 0.3. The directional reward penalizes both speed error and tangential drift: 𝑟dir = ( 0 if 𝑣 ∥ ≤ 0 exp −𝛼𝑣 [PITH_FULL_IMAGE:figures/ful… view at source ↗
Figure 17
Figure 17. Figure 17: Return Curve under different Nucleus Sampling configurations [PITH_FULL_IMAGE:figures/full_fig_p017_17.png] view at source ↗
read the original abstract

Developing controllers capable of completing a wide range of tasks in a natural and life-like manner is a key challenge in enabling practical applications of physics-based character animation. In this work, we introduce Generative Pretrained Controllers (GPC), which leverage tokenization and next-token modeling to create general-purpose, reusable generative controllers from large-scale motion datasets. Our framework utilizes end-to-end reinforcement learning to jointly optimize a "motion vocabulary", modeled via Finite Scalar Quantization (FSQ), along with a corresponding control policy that can map the discrete codes to physics-based controls. After the "codebook" has been learned, the underlying structure of this large vocabulary is modeled by training a GPT-style autoregressive transformer, leading to a powerful generative controller that generates controls for a physically simulated character by performing next-token prediction. Once the generative controller has been trained, we propose a suite of adaptation techniques for finetuning the controller for new downstream tasks. Our proposed framework greatly simplifies the training process compared to previous tokenized methods, and achieves a 99.98% success rate in reproducing a vast corpus of motion clips. The generative controller exhibits a variety of natural emergent behaviors, such as responsive behaviors to perturbations and recovery behaviors after falling. This results in highly robust general purpose controllers for a variety of downstream applications.

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

2 major / 0 minor

Summary. The paper introduces Generative Pretrained Controllers (GPC) that jointly optimize a motion vocabulary via Finite Scalar Quantization (FSQ) and a control policy through end-to-end reinforcement learning on large-scale motion datasets. After codebook learning, a GPT-style autoregressive transformer models the vocabulary structure to generate physics-based controls via next-token prediction; the resulting controller is adapted to downstream tasks and reportedly achieves 99.98% success in reproducing the motion corpus along with emergent behaviors such as perturbation response and recovery.

Significance. If the discrete codes prove to support genuine extrapolation beyond the training distribution, the approach would offer a scalable pretraining route for reusable controllers in physics-based animation, potentially simplifying pipelines relative to prior tokenized methods while enabling robust transfer.

major comments (2)
  1. [Abstract] Abstract: the 99.98% reproduction success rate is presented without any description of held-out motion clips, train/test splits, error bars, or baseline comparisons, which directly undermines the central claim that the learned codes support transferable, general-purpose controllers rather than in-distribution reconstruction.
  2. [Framework description] Framework (joint FSQ+RL optimization): no analysis or ablation is supplied to show that the codes learned under the RL objective form sequences whose statistics are sufficiently regular and hierarchically structured for an autoregressive transformer to extrapolate to unseen motion statistics; the reported reproduction rate is consistent with codes optimized only for task fidelity.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our work. We address each major comment below and will revise the manuscript to improve the presentation of results and add supporting analysis.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the 99.98% reproduction success rate is presented without any description of held-out motion clips, train/test splits, error bars, or baseline comparisons, which directly undermines the central claim that the learned codes support transferable, general-purpose controllers rather than in-distribution reconstruction.

    Authors: We agree that the abstract would benefit from greater clarity on the evaluation. The 99.98% figure measures reproduction success on the motion corpus used during pretraining. The manuscript demonstrates transfer through adaptation to downstream tasks and emergent robustness properties. We will revise the abstract to explicitly note the in-distribution nature of the reproduction metric, reference the train/test protocol and error bars from the experiments section, and highlight baseline comparisons already present in the full paper. This will better contextualize the result without overstating generalization from reproduction alone. revision: yes

  2. Referee: [Framework description] Framework (joint FSQ+RL optimization): no analysis or ablation is supplied to show that the codes learned under the RL objective form sequences whose statistics are sufficiently regular and hierarchically structured for an autoregressive transformer to extrapolate to unseen motion statistics; the reported reproduction rate is consistent with codes optimized only for task fidelity.

    Authors: The end-to-end RL objective is intended to yield codes that are both task-effective and statistically regular enough for autoregressive modeling, which is evidenced by the successful training of the GPT-style transformer and the emergence of perturbation response and recovery behaviors. We acknowledge, however, that dedicated ablations examining code sequence statistics, hierarchical structure, and extrapolation to unseen motion distributions would strengthen the argument that the codes are not merely task-optimized. We will incorporate such analysis and ablations in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation chain is self-contained

full rationale

The described pipeline (end-to-end RL joint optimization of FSQ vocabulary + policy, followed by separate GPT-style transformer training on extracted token sequences, then downstream adaptation) does not reduce any load-bearing claim to its inputs by construction. The 99.98% reproduction metric is an in-distribution evaluation on the training corpus but is not presented as a 'prediction' that is definitionally identical to the fit; the transferability claims rest on the autoregressive modeling step and adaptation techniques, which are independent modeling choices rather than tautological. No self-citations, uniqueness theorems, or ansatzes imported from prior author work are invoked in the provided text to justify core steps. The framework is a standard tokenization + autoregressive pretraining approach and remains externally falsifiable on held-out tasks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Only the abstract is available; no specific free parameters, axioms, or invented entities can be extracted or audited beyond the high-level reliance on standard techniques (FSQ, RL, transformers) whose details and assumptions remain unspecified.

pith-pipeline@v0.9.1-grok · 5771 in / 1304 out tokens · 55540 ms · 2026-06-30T08:01:30.892502+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

296 extracted references · 43 canonical work pages · 9 internal anchors

  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. 2007. doi:10.1145/1188913.1188915

    work page doi:10.1145/1188913.1188915 2007

  2. [2]

    Deciding equivalances among conjunctive aggregate queries

    Sarah Cohen and Werner Nutt and Yehoshua Sagic. Deciding equivalances among conjunctive aggregate queries. 2007. doi:10.1145/1219092.1219093

    work page doi:10.1145/1219092.1219093 2007

  3. [3]

    Special issue: Digital Libraries. 1996

    1996

  4. [6]

    The title of book two. 2008. doi:10.1007/3-540-09237-4

    work page doi:10.1007/3-540-09237-4 2008

  5. [7]

    Asad Z. Spector. Achieving application requirements. Distributed Systems. 1990. doi:10.1145/90417.90738

    work page doi:10.1145/90417.90738 1990

  6. [8]

    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. 1998. doi:10.1007/3-540-65193-4_29

    work page doi:10.1007/3-540-65193-4_29 1998

  7. [9]

    Donald E. Knuth. The Art of Computer Programming, Vol. 1: Fundamental Algorithms (3rd. ed.). 1997

    1997

  8. [10]

    Donald E. Knuth. The Art of Computer Programming. 1998

    1998

  9. [11]

    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

  10. [12]

    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. 2007

    2007

  11. [13]

    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. 2008

    2008

  12. [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. 2009

    2009

  13. [15]

    Predicate Path expressions

    Sten Andler. Predicate Path expressions. Proceedings of the 6th. ACM SIGACT-SIGPLAN symposium on Principles of Programming Languages. 1979. doi:10.1145/567752.567774

    work page doi:10.1145/567752.567774 1979

  14. [16]

    LOGICS of Programs: AXIOMATICS and DESCRIPTIVE POWER

    David Harel. LOGICS of Programs: AXIOMATICS and DESCRIPTIVE POWER. 1978

    1978

  15. [17]

    Anisi , title =

    David A. Anisi , title =

  16. [18]

    Clarkson

    Kenneth L. Clarkson. Algorithms for Closest-Point Problems (Computational Geometry). 1985

    1985

  17. [19]

    Introduction to Bayesian Statistics

    Harry Thornburg. Introduction to Bayesian Statistics. 2001

    2001

  18. [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

  19. [21]

    Stats and Analysis

    Poker-Edge.Com. Stats and Analysis. 2006

    2006

  20. [22]

    A more perfect union

    Barack Obama. A more perfect union. 2008

    2008

  21. [23]

    The fountain of youth

    Joseph Scientist. The fountain of youth. 2009

    2009

  22. [24]

    Solder man

    Dave Novak. Solder man. ACM SIGGRAPH 2003 Video Review on Animation theater Program: Part I - Vol. 145 (July 27--27, 2003). 2003. doi:99.9999/woot07-S422

    2003

  23. [25]

    Interview with Bill Kinder: January 13, 2005

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

    work page doi:10.1145/1057270.1057278 2005

  24. [26]

    The Enabling of Digital Libraries

    Bernard Rous. The Enabling of Digital Libraries. Digital Libraries. 2008

    2008

  25. [28]

    (new) Finding minimum congestion spanning trees , journal =

    Werneck, Renato and Setubal, Jo\. (new) Finding minimum congestion spanning trees , journal =. 2000 , issn =. doi:10.1145/351827.384253 , acmid =

    work page doi:10.1145/351827.384253 2000

  26. [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 =

    work page doi:10.1016/j.inffus.2009.01.002 2009

  27. [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 =

    work page doi:10.1145/1358628.1358946

  28. [32]

    , title =

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

    1999

  29. [33]

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

    1999

  30. [34]

    and Rosenberg, Arnold L

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

    1987

  31. [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 =

  32. [36]

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

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

  33. [37]

    SIGCOMM Comput. Commun. Rev. , year =

  34. [38]

    2004 , isbn =

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

    work page doi:10.1109/icws.2004.64 2004

  35. [39]

    Distributed systems (2nd Ed.) , year =

  36. [40]

    , title =

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

    1986

  37. [41]

    Donald E. Knuth. Seminumerical Algorithms. 1981

    1981

  38. [42]

    E-commerce and cultural values , year =

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

  39. [43]

    E-commerce and cultural values , year =

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

  40. [44]

    Chapter 9 , booktitle =

    Kong, Wei-Chang , editor =. Chapter 9 , booktitle =. 2002 , address =

    2002

  41. [45]

    E-commerce and cultural values , editor =

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

    2003

  42. [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

  43. [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

  44. [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

  45. [49]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    RoHM: Robust Human Motion Reconstruction via Diffusion , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  46. [50]

    Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=

    TMR: Text-to-motion retrieval using contrastive 3D human motion synthesis , author=. Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=

  47. [51]

    ACM Transactions On Graphics (TOG) , volume=

    Deepmimic: Example-guided deep reinforcement learning of physics-based character skills , author=. ACM Transactions On Graphics (TOG) , volume=. 2018 , publisher=

    2018

  48. [52]

    ACM Transactions on Graphics (ToG) , volume=

    Amp: Adversarial motion priors for stylized physics-based character control , author=. ACM Transactions on Graphics (ToG) , volume=. 2021 , publisher=

    2021

  49. [53]

    ACM Transactions On Graphics (TOG) , volume=

    Ase: Large-scale reusable adversarial skill embeddings for physically simulated characters , author=. ACM Transactions On Graphics (TOG) , volume=. 2022 , publisher=

    2022

  50. [54]

    ACM Transactions on Graphics (TOG) , volume=

    Interactive character control with auto-regressive motion diffusion models , author=. ACM Transactions on Graphics (TOG) , volume=. 2024 , publisher=

    2024

  51. [55]

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

    AMASS: Archive of motion capture as surface shapes , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=

  52. [56]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Generating diverse and natural 3d human motions from text , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  53. [57]

    ACM SIGGRAPH 2023 Conference Proceedings , keywords =

    Tessler, Chen and Kasten, Yoni and Guo, Yunrong and Mannor, Shie and Chechik, Gal and Peng, Xue Bin , title =. ACM SIGGRAPH 2023 Conference Proceedings , keywords =. 2023 , isbn =

    2023

  54. [58]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Momask: Generative masked modeling of 3d human motions , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  55. [59]

    ACM Transactions on Graphics (TOG) , volume=

    Object motion guided human motion synthesis , author=. ACM Transactions on Graphics (TOG) , volume=. 2023 , publisher=

    2023

  56. [60]

    ACM Transactions on Graphics (TOG) , volume=

    Robust solving of optical motion capture data by denoising , author=. ACM Transactions on Graphics (TOG) , volume=. 2018 , publisher=

    2018

  57. [61]

    European Conference on Computer Vision , pages=

    COIN: Control-Inpainting Diffusion Prior for Human and Camera Motion Estimation , author=. European Conference on Computer Vision , pages=. 2025 , organization=

    2025

  58. [62]

    ACM SIGGRAPH 2024 Conference Papers , pages=

    Flexible motion in-betweening with diffusion models , author=. ACM SIGGRAPH 2024 Conference Papers , pages=

    2024

  59. [63]

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

    Humor: 3d human motion model for robust pose estimation , author=. Proceedings of the IEEE/CVF international conference on computer vision , pages=

  60. [64]

    The Eleventh International Conference on Learning Representations , year=

    Human Motion Diffusion Model , author=. The Eleventh International Conference on Learning Representations , year=

  61. [65]

    Advances in Neural Information Processing Systems , volume=

    Motion-x: A large-scale 3d expressive whole-body human motion dataset , author=. Advances in Neural Information Processing Systems , volume=

  62. [66]

    Advances in neural information processing systems , volume=

    Decision transformer: Reinforcement learning via sequence modeling , author=. Advances in neural information processing systems , volume=

  63. [67]

    Advances in neural information processing systems , volume=

    Offline reinforcement learning as one big sequence modeling problem , author=. Advances in neural information processing systems , volume=

  64. [68]

    ACM Transactions on Graphics (TOG) , volume=

    Listen, denoise, action! audio-driven motion synthesis with diffusion models , author=. ACM Transactions on Graphics (TOG) , volume=. 2023 , publisher=

    2023

  65. [69]

    International Conference on Learning Representations , year=

    Should I Run Offline Reinforcement Learning or Behavioral Cloning? , author=. International Conference on Learning Representations , year=

  66. [70]

    Transactions on Machine Learning Research , issn=

    Soft Diffusion: Score Matching with General Corruptions , author=. Transactions on Machine Learning Research , issn=. 2023 , url=

    2023

  67. [71]

    arXiv preprint arXiv:2411.02780 , year=

    How much is a noisy image worth? Data Scaling Laws for Ambient Diffusion , author=. arXiv preprint arXiv:2411.02780 , year=

    work page arXiv

  68. [72]

    Advances in Neural Information Processing Systems , volume=

    Ambient diffusion: Learning clean distributions from corrupted data , author=. Advances in Neural Information Processing Systems , volume=

  69. [73]

    Forty-first International Conference on Machine Learning , year=

    Consistent Diffusion Meets Tweedie: Training Exact Ambient Diffusion Models with Noisy Data , author=. Forty-first International Conference on Machine Learning , year=

  70. [74]

    The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=

    Learning Diffusion Priors from Observations by Expectation Maximization , author=. The Thirty-eighth Annual Conference on Neural Information Processing Systems , year=

  71. [75]

    Proceedings of the AAAI Conference on Artificial Intelligence , volume=

    Leveraging contaminated datasets to learn clean-data distribution with purified generative adversarial networks , author=. Proceedings of the AAAI Conference on Artificial Intelligence , volume=

  72. [76]

    Forty-first International Conference on Machine Learning , year=

    Contamination-Resilient Anomaly Detection via Adversarial Learning on Partially-Observed Normal and Anomalous Data , author=. Forty-first International Conference on Machine Learning , year=

  73. [77]

    Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

    Noisier2noise: Learning to denoise from unpaired noisy data , author=. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages=

  74. [78]

    International Conference on Machine Learning , pages=

    Noise2Noise: Learning image restoration without clean data , author=. International Conference on Machine Learning , pages=. 2018 , organization=

    2018

  75. [79]

    Dimakis , booktitle=

    Ashish Bora and Eric Price and Alexandros G. Dimakis , booktitle=. Ambient. 2018 , url=

    2018

  76. [80]

    Motion in Games: First International Workshop, MIG 2008, Utrecht, The Netherlands, June 14-17, 2008

    Automatic estimation of skeletal motion from optical motion capture data , author=. Motion in Games: First International Workshop, MIG 2008, Utrecht, The Netherlands, June 14-17, 2008. Revised Papers 1 , pages=. 2008 , organization=

    2008

  77. [81]

    ACM SIGGRAPH 2010 papers , pages=

    Sampling-based contact-rich motion control , author=. ACM SIGGRAPH 2010 papers , pages=. 2010 , publisher=

    2010

  78. [82]

    Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=

    Perpetual humanoid control for real-time simulated avatars , author=. Proceedings of the IEEE/CVF International Conference on Computer Vision , pages=

  79. [83]

    ACM Transactions on Graphics (TOG) , volume=

    Moconvq: Unified physics-based motion control via scalable discrete representations , author=. ACM Transactions on Graphics (TOG) , volume=. 2024 , publisher=

    2024

  80. [84]

    2024 , journal=

    Tessler, Chen and Guo, Yunrong and Nabati, Ofir and Chechik, Gal and Peng, Xue Bin , title =. 2024 , journal=

    2024

Showing first 80 references.