Pith. sign in

REVIEW 3 major objections 5 minor 25 references

From Text to Pose to Image: Improving Diffusion Model Control and Quality

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read Two new models enable the first generative text-to-pose-to-image framework for diffusion-based human pose control, improving pose fidelity and aesthetics over the prior adapter.

desk verdict Useful pipeline, honest limits, but the central T2P claim rests on a self-trained metric over 100 pairs. read the letter →

arxiv 2411.12872 v2 pith:TZKP6Z6G submitted 2024-11-19 cs.CV cs.AIcs.LG

Clément Bonnet , Ariel N. Lee , Franck Wertel , Antoine Tamano , Tanguy Cizain , Pablo Ducru This is my paper · ORCID
classification cs.CVcs.AIcs.LG
keywords text-to-posediffusionmodelsposecontroladaptercontrastivelanguage-posepretrainingGaussianmixturetransformertempereddistributionsamplingimagegeneration
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 aims to close the loop between text prompts and posed human images by splitting the task in two. It introduces T2P, an autoregressive transformer that generates a skeleton of body, face, and hand keypoints directly from a text description, and a new pose adapter that conditions a diffusion model on that fuller skeleton. Together they form, the paper argues, the first generative text-to-pose-to-image pipeline, giving users control over human pose from text alone rather than by retrieving poses from a database. The paper reports that T2P beats nearest-neighbor pose retrieval on 78 percent of COCO caption-posed pairs, and that its adapter beats the previous Tencent SDXL adapter on 70 percent of aesthetic-score comparisons and 76 percent on human-preference scores. A reader would care because this is a concrete route to controllable human-figure generation for SDXL-class diffusion models.

What carries the argument

The load-bearing objects are three: CLaPP, a contrastive text-pose metric trained on 500k JourneyDB caption-pose pairs that scores how well a prompt matches a skeleton and serves as the training signal and evaluation metric for T2P; T2P, an autoregressive decoder-only transformer that predicts pose keypoints as a continuous sequence, parameterized as a mixture of six Gaussians with a binary existence head, and sampled at inference via tempered distribution sampling; and a pose adapter built on the T2I-Adapter architecture but trained with all 128 whole-body keypoints. The tempered sampling transform (Theorem 1) is what keeps the autoregressive GMM outputs on-distribution at inference; without it, generated poses drift into noise.

What would settle it

Take the 100 COCO-2017 caption-pose pairs used in the benchmark, generate a pose with T2P and retrieve one with KNN, then have human raters blind to source pick the better-matched pose; if human agreement with CLaPP is not significantly above chance, the reported 78 percent win rate does not establish that T2P generalizes to new descriptions.

Watch

Extended reading notes

Core claim

The central discovery is that pose control for text-to-image diffusion can be factored into two learnable stages. First, a decoder-only transformer over ordered keypoints predicts the next (x,y) coordinate, using a Gaussian mixture model for the continuous distribution and a binary classifier for keypoint existence; cross-attention on CLIP text features makes the generation prompt-conditioned. Because sampling from the raw mixture at temperature 1 diverges, the paper defines a tempered distribution transform and gives a Monte Carlo sampling scheme based on softmax weights $e^{(1/T-1)\ln p(x_i)}$. Second, a new adapter conditioned on 18 body plus 42 hand plus 68 face keypoints is trained on high-quality images; the paper reports that this adapter, unlike the Tencent baseline, preserves hand and face structure and yields better aesthetic and human-preference scores. The authors conclude that, together, these two models constitute the first generative text-to-pose-to-image framework for higher pose control in diffusion models.

Load-bearing premise

The claim that T2P generalizes rests on CLaPP's text-pose similarity scores being a faithful measure of real alignment; the paper itself notes CLaPP may be biased for out-of-distribution captions or poses and that CLIP backbones may be pose-agnostic.

Editorial extensions

If this is right

  • Users can go from a caption to a pose skeleton to a generated image without supplying a reference photo, because T2P synthesizes poses instead of retrieving them from a dataset.
  • The fuller 128-keypoint conditioning (body, face, and hands) lets the new adapter preserve hand and facial structure, addressing a known weakness of the Tencent adapter.
  • Because the pose is an explicit intermediate, a user can edit the skeleton and regenerate with the same seed, turning pixel-level editing into pose-level editing.
  • The reported win rates suggest that both stages together offer better pose fidelity and aesthetics than the previous adapter when evaluated on the paper's COCO-Pose benchmark.

Reading between the lines

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

  • A natural next step, not pursued in the paper, would be to replace the image-based CLaPP metric with a sequence-based contrastive model over (x,y) coordinates, which the paper itself notes could be lower-dimensional and less pose-agnostic.
  • The tempered sampling theorem is not specific to poses: any autoregressive generative model whose output is a mixture density could use the same softmax reweighting scheme to sharpen or flatten samples at inference, so the algorithm may transfer to other continuous-sequence generators.
  • The factorized text-to-pose-to-image design could be chained with depth or segmentation adapters to build a fuller semantic-to-image pipeline where humans, objects, and scenes are each controlled by an explicit intermediate representation.
  • Because the paper reports that T2P lacks diversity at low temperatures, a testable extension is to combine temperature scheduling with latent-space pose editing to recover diversity without sacrificing the precision that makes the generated poses usable.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a complete text-to-pose-to-image pipeline for controlling human pose in text-to-image diffusion models. It introduces CLaPP, a contrastive text-pose metric trained on 500k JourneyDB pairs; T2P, an autoregressive transformer that predicts 2D whole-body pose keypoints (body, face, hands) from CLIP text features using a Gaussian mixture model and a binary existence classifier; a tempered-distribution sampling scheme for GMMs; and a new SDXL pose adapter trained on whole-body keypoints. The authors report a 78% CLaPP win rate for T2P over KNN retrieval on 100 COCO pairs and 70%/76% win rates for their adapter over the SDXL-Tencent adapter on aesthetic score and HPS v2, plus a human preference test. The central claim is that these components constitute, for the first time, a generative text-to-pose-to-image framework with improved pose control.

Significance. If the empirical claims were substantiated, the factorization of text-to-image generation through an intermediate pose representation would be practically useful, and the tempered-sampling identity for GMMs is a clean and reusable theoretical contribution. The release of code and models is a concrete strength. However, the evidence presented is not yet sufficient to support the paper's central novelty and state-of-the-art claims: the main text-to-pose evaluation relies on a metric trained by the same authors on the same data distribution, and the adapter comparison covers only one baseline. The paper's own Appendix A.5 acknowledges several of these weaknesses.

major comments (3)
  1. [§2.5, Fig. 2] The 78% win rate for T2P over KNN is computed exclusively with CLaPP, a contrastive metric trained by the same authors on 500k JourneyDB pairs, and Appendix A.5 explicitly concedes that CLaPP may be biased for out-of-distribution captions or poses and that the CLIP backbone may be pose-agnostic. Since T2P and CLaPP are trained on the same JourneyDB distribution while the benchmark captions come from COCO, this is a self-reported alignment score rather than an independent measure of text-pose compatibility. With only 100 pairs, the confidence interval is also wide. I would require at least one independent evaluation, such as human judgments of text-pose alignment or a pose-retrieval test on held-out captions, and a larger benchmark before accepting the central claim that T2P generalizes beyond the training distribution.
  2. [§3.2, Fig. 3] The adapter evaluation compares only against the SDXL-Tencent adapter, yet the paper describes the result as state-of-the-art and claims in §4 that the adapter 'guarantees both high fidelity to the pose and high image aesthetics.' No comparison is made to ControlNet/OpenPose or other pose-conditioning methods, and no quantitative pose fidelity metric (e.g., PCK or OKS computed with DWPose) is reported; Figure 8 provides only visual examples. Figure 3(c) reports human preference percentages without stating the number of raters, the number of judgments, or the exact values. These gaps are load-bearing because the adapter's superiority is one of the paper's two main contributions.
  3. [§2.3, §A.5] The paper introduces T2P as a generative model and uses it to claim a first-of-its-kind text-to-pose-to-image framework, but Appendix A.5 states that T2P 'lacks diversity and is really just a reflection of the data it was trained on,' and no quantitative diversity or prompt-conditional fidelity metrics are reported anywhere. The only quantitative comparison is against KNN retrieval, with no positioning against the existing text-to-motion or pose-generation literature. The 'first of its kind' claim therefore needs both stronger evidence and a more thorough literature comparison.
minor comments (5)
  1. [§2.2] The fixed ordering of 128 keypoints is not fully specified: the paper should state how missing keypoints (e.g., face points for a back view or absent hands) are represented in the input sequence and how they interact with the binary existence classifier.
  2. [§3.2, Fig. 3] The error bars in Figures 3(a) and 3(b) and the human-preference percentages in Figure 3(c) are not accompanied by the underlying counts or a description of how raters were instructed; please add these details.
  3. [§A.3] Theorem 1 should explicitly state the integrability and normalizability assumptions on p, and the mode-selection statement should be qualified for distributions with multiple equal modes; as written, the Dirac limit is informal.
  4. [§2.1] The CLaPP training procedure is under-specified: the loss function, negative sampling strategy, batch size, and the architecture of the 'CLaPP layers' are not described, which makes the metric difficult to reproduce or audit.
  5. [Throughout] Please fix typographical errors in the figures and references, including 'T wo' in Figure 5, 'T encent' in Figure 8, and 'V on' in reference [16].

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central T2P and adapter results are empirical evaluations against external benchmarks, and no derivation reduces to its own inputs.

full rationale

The paper's central claims are the T2P generative model and the pose adapter. The T2P win rate over KNN is measured with CLaPP, a contrastive metric introduced by the same authors, and Appendix A.5 concedes that CLaPP may be biased for out-of-distribution captions or poses and that CLIP backbones may be pose-agnostic. However, this is an evaluation-validity concern, not circularity: T2P is trained by maximum likelihood on next-pose GMM parameters (Section 2.3), not by optimizing CLaPP, so the 78% result is not forced by construction. There is no equation in the paper that equates the predicted pose quality with CLaPP's training objective, nor any fitted parameter renamed as a prediction. The adapter is evaluated with external aesthetic scores, HPS v2, and human preference annotations, which are independent of the model's training data and objectives. The tempered-distribution 'theorem' states properties that follow algebraically from its own definition and is presented as such, not as an empirical discovery. No load-bearing self-citation or uniqueness theorem is invoked. Accordingly, the derivation chain is self-contained, and the appropriate finding is no significant circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The paper's central T2P and adapter claims rest on pose annotations from DWPose, CLIP representations, and the JourneyDB dataset, all treated as trusted inputs. The tempered-sampling result depends on the validity of importance-sampling approximation at finite N. The hyperparameters are chosen empirically without a clean validation protocol.

free parameters (4)
  • Number of GMM mixture components = 6
    Section 2.3: 'a mixture of 6 different Gaussian mixtures ... worked best for our dataset'. Chosen without a reported validation split.
  • Number of transformer layers = 4
    Section 2.3: 'a transformer with 4 layers worked best for our dataset'. Hyperparameter chosen on the training set.
  • Inference temperature T = not numerically specified in text
    Section 2.4 and Figure 7: sampling temperature is lowered to improve precision at the expense of diversity; no reported tuned value or sensitivity analysis.
  • Number of Monte Carlo samples N for tempered sampling = 10,000 in toy example; unspecified for T2P
    Figure 6 uses N=10,000; the actual N used for T2P inference is not stated, and the approximation is presented as a theorem without error bounds.
assumptions (5)
  • domain assumption DWPose extracts accurate body, face, and hand keypoints from images
    Used to annotate JourneyDB for CLaPP and T2P training (Sections 2.1 and 2.3); errors in pose extraction would propagate.
  • ad hoc to paper The fixed ordering of 128 keypoints (18 body, 68 face, 42 hands) is a valid sequence representation for autoregressive generation
    Section 2.2: poses are modeled as ordered sequences; this ordering is arbitrary and affects the model's ability to capture spatial relations.
  • standard math Tempered distribution sampling via importance sampling converges adequately with finite N
    Theorem 1 and Proof 1: the sampling scheme is justified by importance sampling, but no finite-sample error bound or convergence diagnostic is given.
  • domain assumption CLIP text and image embeddings are informative for pose semantics
    CLaPP uses CLIP as a backbone (Section 2.1); the paper itself notes CLIP may be pose-agnostic (Section A.5).
  • domain assumption JourneyDB prompts and images are representative of the target deployment distribution
    T2P and CLaPP are trained on JourneyDB (Section 2.3); if the distribution differs from real user prompts, performance may degrade.

how reviews work

0 comments
Cite this review

Pith. "Pith review of From Text to Pose to Image: Improving Diffusion Model Control and Quality." pith.science (2026). https://pith.science/paper/TZKP6Z6G

@misc{pith2026241112872,
  author       = {Pith},
  title        = {Pith review of: From Text to Pose to Image: Improving Diffusion Model Control and Quality},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TZKP6Z6G}},
  note         = {Machine review of arXiv:2411.12872}
}
read the original abstract

In the last two years, text-to-image diffusion models have become extremely popular. As their quality and usage increase, a major concern has been the need for better output control. In addition to prompt engineering, one effective method to improve the controllability of diffusion models has been to condition them on additional modalities such as image style, depth map, or keypoints. This forms the basis of ControlNets or Adapters. When attempting to apply these methods to control human poses in outputs of text-to-image diffusion models, two main challenges have arisen. The first challenge is generating poses following a wide range of semantic text descriptions, for which previous methods involved searching for a pose within a dataset of (caption, pose) pairs. The second challenge is conditioning image generation on a specified pose while keeping both high aesthetic and high pose fidelity. In this article, we fix these two main issues by introducing a text-to-pose (T2P) generative model alongside a new sampling algorithm, and a new pose adapter that incorporates more pose keypoints for higher pose fidelity. Together, these two new state-of-the-art models enable, for the first time, a generative text-to-pose-to-image framework for higher pose control in diffusion models. We release all models and the code used for the experiments at https://github.com/clement-bonnet/text-to-pose.

Figures

Figures reproduced from arXiv: 2411.12872 by the authors.

Figure 1
Figure 1. Text-to-Pose transformer architecture. A pose is defined as an ordered sequence of key points, i.e. (x, y) coordinates of points in the image that correspond to e.g. the right thumb, the left shoulder, the nose, etc. Given this, it makes sense to design a sequence model that can embed the whole pose conditioned on text features. For this, we use a decoder-only transformer architecture [16] (see figure 1) to auto-reg… view at source ↗
Figure 2
Figure 2. CLaPP scores with 95% confidence intervals. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Performance of pose-conditioned image generation for the Tencent adapter and ours. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Text-to-pose-to-image framework. A.2 CLaPP metric Our contrastive model, called CLaPP, can predict compatibility scores between a prompt and a human pose [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: CLaPP scores on 5 poses and corresponding captions from the COCO dataset. The scores [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 8
Figure 8. Figure 8: figure 8 [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 6
Figure 6. Figure 6: Tempered distribution sampling of Gaussian Mixture Model. [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Effect of the temperature on GMM tempered sampling for text-to-pose generation. Prompt: [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]
Figure 8
Figure 8. Figure 8: Pose-conditioned image generation using the SDXL-Tencent adapter, and ours (which [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 7 canonical work pages

  1. [1]

    Cover and P

    T. Cover and P. Hart. Nearest neighbor pattern classification.IEEE Transactions on Information Theory, 13(1):21–27, 1967

    1967

  2. [2]

    An image is worth one word: Personalizing text-to-image generation using textual inversion

    Rinon Gal, Yuval Alaluf, Yuval Atzmon, Or Patashnik, Amit H Bermano, Gal Chechik, and Daniel Cohen-Or. An image is worth one word: Personalizing text-to-image generation using textual inversion. arXiv preprint arXiv:2208.01618, 2022

    arXiv 2022

  3. [3]

    Distilling the Knowledge in a Neural Network,

    Geoffrey Hinton, Oriol Vinyals, and Jeff Dean. Distilling the Knowledge in a Neural Network,

    work page

  4. [4]

    Denoising diffusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffusion probabilistic models. In H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, editors, Advances in Neural Information Processing Systems, volume 33, pages 6840–6851. Curran Associates, Inc., 2020

    work page 2020

  5. [5]

    Elucidating the Design Space of Diffusion-Based Generative Models, 2022

    Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine. Elucidating the Design Space of Diffusion-Based Generative Models, 2022. URL http://arxiv.org/abs/2206.00364. arXiv:2206.00364 [cs, stat]

    arXiv 2022

  6. [6]

    Lawrence Zitnick, and Piotr Dollár

    Tsung-Yi Lin, Michael Maire, Serge Belongie, Lubomir Bourdev, Ross Girshick, James Hays, Pietro Perona, Deva Ramanan, C. Lawrence Zitnick, and Piotr Dollár. Microsoft COCO: Com- mon Objects in Context, 2015. URL http://arxiv.org/abs/1405.0312. arXiv:1405.0312 [cs]

    arXiv 2015

  7. [7]

    Pose Guided Person Image Generation

    Liqian Ma, Xu Jia, Qianru Sun, Bernt Schiele, Tinne Tuytelaars, and Luc Van Gool. Pose Guided Person Image Generation. In Advances in Neural Information Processing Systems , volume 30. Curran Associates, Inc., 2017

    work page 2017

  8. [8]

    SDEdit: Guided image synthesis and editing with stochastic differential equations

    Chenlin Meng, Yutong He, Yang Song, Jiaming Song, Jiajun Wu, Jun-Yan Zhu, and Stefano Ermon. SDEdit: Guided image synthesis and editing with stochastic differential equations. In International Conference on Learning Representations, 2022

    2022

Show all 25 references
  1. [9]

    T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models, 2023

    Chong Mou, Xintao Wang, Liangbin Xie, Yanze Wu, Jian Zhang, Zhongang Qi, Ying Shan, and Xiaohu Qie. T2I-Adapter: Learning Adapters to Dig out More Controllable Ability for Text-to-Image Diffusion Models, 2023. URL http://arxiv.org/abs/2302.08453. arXiv:2302.08453 [cs]

    2023 arXiv

  2. [10]

    Learning transferable visual models from natural language supervision

    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 supervision. In International Conference on Machine Learning, pa...

    2021

  3. [11]

    Hierarchical Text-Conditional Image Generation with CLIP Latents, 2022

    Aditya Ramesh, Prafulla Dhariwal, Alex Nichol, Casey Chu, and Mark Chen. Hierarchical Text-Conditional Image Generation with CLIP Latents, 2022. URL http://arxiv.org/abs/ 2204.06125. arXiv:2204.06125 [cs]

    2022 arXiv

  4. [12]

    High- resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Björn Ommer. High- resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 10684–10695, 2022

    2022

  5. [13]

    Deep Un- supervised Learning using Nonequilibrium Thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep Un- supervised Learning using Nonequilibrium Thermodynamics. In Proceedings of the 32nd International Conference on Machine Learning, pages 2256–2265. PMLR, 2015. ISSN: 1938- 7228

    2015

  6. [14]

    Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole

    Yang Song, Jascha Sohl-Dickstein, Diederik P. Kingma, Abhishek Kumar, Stefano Ermon, and Ben Poole. Score-Based Generative Modeling through Stochastic Differential Equations, 2021. URL http://arxiv.org/abs/2011.13456. arXiv:2011.13456 [cs, stat]

    2021 arXiv

  7. [15]

    Journeydb: A benchmark for generative image understanding, 2023

    Keqiang Sun, Junting Pan, Yuying Ge, Hao Li, Haodong Duan, Xiaoshi Wu, Renrui Zhang, Aojun Zhou, Zipeng Qin, Yi Wang, Jifeng Dai, Yu Qiao, Limin Wang, and Hongsheng Li. Journeydb: A benchmark for generative image understanding, 2023. 5

    2023

  8. [16]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polosukhin. Attention is all you need. In I. Guyon, U. V on Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors,Advances in Neural I...

    2017

  9. [17]

    Awadallah

    Chi Wang, Susan Xueqing Liu, and Ahmed H. Awadallah. Cost-Effective Hyperparameter Optimization for Large Language Model Generation Inference, 2023. URL http://arxiv. org/abs/2303.04673. arXiv:2303.04673 [cs]

    2023 arXiv

  10. [18]

    Contextual Temperature for Language Modeling, 2020

    Pei-Hsin Wang, Sheng-Iou Hsieh, Shih-Chieh Chang, Yu-Ting Chen, Jia-Yu Pan, Wei Wei, and Da-Chang Juan. Contextual Temperature for Language Modeling, 2020. URL http: //arxiv.org/abs/2012.13575. arXiv:2012.13575 [cs]

    2020 arXiv

  11. [19]

    Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis

    Xiaoshi Wu, Yiming Hao, Keqiang Sun, Yixiong Chen, Feng Zhu, Rui Zhao, and Hongsheng Li. Human preference score v2: A solid benchmark for evaluating human preferences of text-to-image synthesis. arXiv preprint arXiv:2306.09341, 2023

    2023 arXiv

  12. [20]

    Tackling the Generative Learning Trilemma with Denoising Diffusion GANs, 2022

    Zhisheng Xiao, Karsten Kreis, and Arash Vahdat. Tackling the Generative Learning Trilemma with Denoising Diffusion GANs, 2022. URL http://arxiv.org/abs/2112.07804. arXiv:2112.07804 [cs, stat]

    2022 arXiv

  13. [21]

    Effective whole-body pose estimation with two-stages distillation

    Zhendong Yang, Ailing Zeng, Chun Yuan, and Yu Li. Effective whole-body pose estimation with two-stages distillation. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 4210–4220, 2023

    2023

  14. [22]

    IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models, 2023

    Hu Ye, Jun Zhang, Sibo Liu, Xiao Han, and Wei Yang. IP-Adapter: Text Compatible Image Prompt Adapter for Text-to-Image Diffusion Models, 2023. URL http://arxiv.org/abs/ 2308.06721. arXiv:2308.06721 [cs]

    2023 arXiv

  15. [23]

    Shihao Zhao, Dongdong Chen, Yen-Chun Chen, Jianmin Bao, Shaozhe Hao, Lu Yuan, and Kwan-Yee K. Wong. Uni-ControlNet: All-in-One Control to Text-to-Image Diffusion Models,

  16. [2015]

    arXiv:1503.02531 [cs, stat]

    URL http://arxiv.org/abs/1503.02531. arXiv:1503.02531 [cs, stat]

    arXiv

  17. [2023]

    A lady holding a smartphone at a cafe

    URL http://arxiv.org/abs/2305.16322. arXiv:2305.16322 [cs]. 6 A Appendix A.1 Text-to-pose-to-image framework We propose factorizing text-to-image generation into semantics generation (human poses) and then semantics-conditioned image generation (see figure 4). This allows for ...

    arXiv

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.