Pith. sign in

REVIEW 4 major objections 6 minor 30 references

A Framework For Image Synthesis Using Supervised Contrastive Learning

T0 review · 4 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Label-guided supervised contrastive learning in both pretraining and GAN phases cuts COCO FID by up to 30.1% across four text-to-image GANs.

desk verdict A clean, incremental idea for T2I GANs whose reported gains are plausible but whose empirical basis is undermined by comparing against official checkpoints rather than re-trained baselines; deserves review but needs major revision. read the letter →

arxiv 2412.03957 v1 pith:W6UIF5HQ submitted 2024-12-05 cs.CV cs.AI

Yibin Liu , Jianyu Zhang , Li Zhang , Shijian Li , Gang Pan This is my paper · ORCID
classification cs.CVcs.AI
keywords text-to-imagegenerationGANsupervisedcontrastivelearningsemanticlabelguidanceimage-textrepresentationmulti-labelCUBdatasetCOCO
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

Typical text-to-image GANs align an image with its caption but ignore the semantic structure among images that share a category label. This paper proposes adding supervised contrastive losses, driven by image labels, to both the representation-pretraining phase and the GAN training phase, so that same-label image and text embeddings are pulled together while different-label pairs are pushed apart. On the COCO dataset the framework lowers FID by 30.1%, 27.3%, 16.2%, and 17.1% for AttnGAN, DM-GAN, SSA-GAN, and GALIP respectively, with smaller but consistent gains on CUB measured by IS and FID. The result is a training-time collaboration that improves generation quality, especially on complex multi-object scenes, without changing the base model architecture.

What carries the argument

The load-bearing object is the label-guided supervised contrastive loss $$\mathcal{L}_{\mathrm{sup}}(u_i,w)=-\frac{1}{|P_s(i)|}\sum_{p\in P_s(i)}\log\frac{\exp(\mathrm{sim}(u_i,w_p)/\tau)}{\sum_{j\ne i}\exp(\mathrm{sim}(u_i,w_j)/\tau)},$$ where $\mathrm{sim}$ is cosine similarity and $P_s(i)$ collects indices whose labels match $u_i$'s label (or whose label sets intersect, in the multi-label COCO case). Around this loss the framework builds a data-sampling strategy that guarantees every sample has a same-label partner, and twin parameter-sharing encoder branches that produce quadruples of image and text representations for both the pretraining phase and the GAN phase. In the GAN phase the same loss is applied to fake images and their captions, so the generator is rewarded for emitting images that cluster with other images of the same category.

What would settle it

Train the framework on COCO with randomly permuted label sets, so that positives are assigned without regard to true semantics; if FID still improves by the reported margin over the unmodified baseline, then label-driven clustering is not the operative cause.

Watch

Extended reading notes

Core claim

The paper's central claim is that inner-modal semantic correspondence—the fact that images with the same label share visual patterns—is an underused signal in text-to-image GANs, and that supervised contrastive learning can convert that signal into better generation. The framework samples, for each image-text pair, a partner pair sharing at least one label, forms two parameter-sharing contrast branches, and adds three supervised contrastive losses ($\mathcal{L}_{\mathrm{sup}}^{\mathrm{img}}$, $\mathcal{L}_{\mathrm{sup}}^{\mathrm{txt}}$, $\mathcal{L}_{\mathrm{sup}}^{\mathrm{i2t}}$) during pretraining, plus image-image and image-text versions during GAN training. With single-label positives on CUB and multi-label positives on COCO, the authors report that AttnGAN's IS rises from 4.36 to 4.61 and its COCO FID falls from 33.10 to 23.14, and every tested baseline improves on both datasets; the largest relative gains appear on the more complex COCO dataset.

Load-bearing premise

The framework assumes that every image has accurate, complete category labels and that label overlap is a trustworthy signal of semantic similarity, so noisy, missing, or overly coarse labels would create false positive pairs and could degrade rather than improve generation.

Editorial extensions

If this is right

  • The framework is architecture-agnostic: applying it to multi-stage GANs (AttnGAN, DM-GAN) and one-stage GANs (SSA-GAN, GALIP) improves FID on COCO and CUB in every case.
  • The ablation attributes the gain to all three components: improved pretrained encoders alone take AttnGAN's COCO FID from 33.10 to 26.90, and adding the GAN-phase image-image loss yields the largest additional drop.
  • The relative benefit is larger on multi-object COCO (up to 30.1% FID reduction) than on single-object CUB, consistent with the claim that label structure is more informative when scenes are complex.
  • Because the pretrained encoders are frozen in the GAN phase, the additional machinery is limited to the two-branch sampling and contrastive losses on generated quadruples.
  • Using the image-image and image-text supervised contrastive losses together gives the best overall result, although adding the image-text loss alone can slightly worsen CUB FID relative to image-image alone.

Reading between the lines

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

  • A direct extension the authors leave to future work is to replace human labels with nouns extracted from captions; the multi-label contrastive loss could absorb such pseudo-labels, and the method's success would then depend on the precision of the extracted positives.
  • The same loss could likely be inserted into diffusion-based text-to-image pipelines, since it operates on encoder outputs and generated images rather than on any GAN-specific objective; the authors mention diffusion only as a possible future direction.
  • The framework's clustering pressure may trade fidelity for diversity: as the authors note, adding the image-text loss can lower CUB FID, and a dedicated intra-class diversity metric would reveal whether same-label positives reduce variety within a category.
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 / 6 minor

Summary. The paper proposes a framework that adds label-guided supervised contrastive learning to text-to-image GANs in both the representation pretraining phase and the GAN training phase. The method samples same-label image-text pairs, computes supervised contrastive losses for image-image, text-text, and image-text associations (single-label for CUB, multi-label for COCO), and adds these to the original objectives with weights lambda1 and lambda2. The framework is applied to AttnGAN, DM-GAN, SSA-GAN, and GALIP, and the authors report improved IS and FID on CUB and COCO, with the headline result being a 30.1% FID improvement for AttnGAN on COCO. An ablation on AttnGAN decomposes the contribution of the pretraining component, the image-image loss, and the image-text loss.

Significance. If the empirical claims hold, the framework is a simple and general recipe for injecting label information into T2I GANs, and the choice of four baselines spanning multi-stage and one-stage architectures is appropriate. The loss definitions in Eqs. (1)-(9) are coherent, and the two-branch parameter-sharing design is clearly described. The paper also makes falsifiable quantitative predictions and includes an ablation study, which are strengths. However, the empirical verification is currently weakened by the use of public checkpoints as baselines rather than identically re-trained baselines, by the absence of variance or multiple-seed reporting for FID, and by the lack of released code or checkpoints. These issues bear directly on the central claim of large FID improvements, so the contribution is not yet fully established.

major comments (4)
  1. [§4.1, Table 1] The note under Table 1 states that baseline results marked * are 'obtained from publicly released pre-trained models by the authors', while the '+SCL' results come from models newly trained by the authors. This mixes evaluation conditions: differences in data preprocessing, training epochs, random seeds, and evaluation code can produce FID changes of the magnitude claimed (e.g., 33.10 to 23.14 for AttnGAN on COCO). The central claim that the framework improves FID by 30.1% is therefore not supported unless each baseline is re-trained by the authors under the same pipeline (same data, same seeds, same evaluation script) and those re-trained baseline numbers are reported in Table 1.
  2. [§4.1, Table 1] No variance or number of seeds is reported for any FID score, and the IS values are reported with standard deviation but it is not stated over how many runs or seeds. GAN training is stochastic, and FID is known to vary across runs. Without at least three seeds and mean±std for both IS and FID, the reported improvements cannot be distinguished from run-to-run noise. This is particularly important for the smaller claimed gains (e.g., GALIP CUB FID 10.08 to 9.90).
  3. [§4.3, Table 2] The ablation shows that the pretraining component alone (ID 2) reduces AttnGAN COCO FID from 33.10 to 26.90, an 18.7% improvement, while the full method reaches 23.14. Because the ID 1 baseline is a public checkpoint rather than a model re-trained with the authors' data pipeline and training schedule, the improvement attributed to the GAN-phase losses (ID 3 vs ID 2, and ID 5 vs ID 2) is measured relative to a baseline that was not subject to the same pipeline. The ablation should include an AttnGAN baseline re-trained by the authors as ID 1, with all configurations trained for the same number of epochs, to isolate the effect of the proposed losses.
  4. [§4 (Implementation Details)] The values of lambda1 and lambda2 are set separately for each dataset and each baseline, and no sensitivity analysis is reported. Since these hyperparameters control the strength of the proposed losses, the reported gains could be affected by the choice of their values. Please provide at least a small sensitivity study (e.g., varying lambda2 for one baseline per dataset) or justify the chosen values by a fixed rule.
minor comments (6)
  1. [§3.1, Eq. (1)] The notation fYp, eYi, and eY is not explained; the symbols appear to mix a tilde or accent with Y. Please define Y_i, Y'_i, and the concatenated label set explicitly.
  2. [§3.1, multi-label loss] In the definition of Pm(i), the condition should be that the labels have non-empty intersection, but the text writes 'fYp ∩ eYi' without explaining the intersection for the multi-label case. Please clarify the notation.
  3. [§4.3] The statement that FID 'measures the KL divergence between the real images and generated images' is incorrect; FID computes the Fréchet distance between feature distributions. Please correct this explanation.
  4. [§4.4] The claim that no existing approach leverages labels like the proposed framework is contradicted by the UniCL and cross-entropy comparisons in the same section; please phrase the novelty claim more carefully.
  5. [References] Reference [23] is cited for FID but points to 'Coulomb GANs' (Unterthiner et al.); the standard FID reference is Heusel et al., 'GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium'. Please fix the citation.
  6. [Throughout] There are several typos, including 'Lorgin' in Eq. (5), 'vice verca' in Section 2.1, and 'noval' in Section 1. Please correct them.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the claimed improvements are empirical results measured with external FID/IS metrics, and the supervised contrastive losses are standard objectives that do not encode the evaluation metric.

full rationale

The paper makes no derivation from first principles; its central claim is an empirical comparison of T2I GAN baselines with and without an added supervised contrastive objective. The proposed losses (Eqs. 1-9) are defined from image/text representations and semantic labels, while the reported gains are measured with Inception Score and Fréchet Inception Distance computed on generated images against external datasets. No fitted parameter is renamed as a prediction, and no equation reduces to the evaluation metric by construction. The baselines are prior published methods cited normally, not self-citations, and the paper's own ablation study (Table 2) separately quantifies the effect of each loss component. The principal weakness is experimental comparison fairness: Table 1 marks baseline numbers with an asterisk as taken from publicly released pretrained models, so the baselines may not have been retrained under identical settings, and no multiple-seed variance is reported for FID. That is an empirical rigor concern about whether the improvement is partly attributable to retraining, not circularity, because the claimed result is not forced by the method's definitions or by any self-citation chain. The acknowledged limitation that the framework is only demonstrated on datasets with detailed label annotations further confirms that labels are an external input rather than an artifact of the evaluation. Overall, no step in the paper's argument reduces to its own inputs, so the appropriate finding is no significant circularity.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The central claim rests on two domain assumptions (label quality and usefulness; frozen-encoder feature quality) and on hand-tuned hyperparameters (λ1, λ2, τ). No new physical or mathematical entities are introduced.

free parameters (3)
  • λ1 (supervised contrastive weight, pretraining) = 0.5 for CUB, 0.05 for COCO
    Set by hand per dataset; controls the balance between the original alignment loss and the supervised contrastive loss in Eq. (5).
  • λ2 (supervised contrastive weight, GAN phase) = CUB: 5, 2.5, 0.2, 0.15; COCO: 2.5, 2.5, 0.1, 0.15 for AttnGAN, DM-GAN, SSA-GAN, GALIP
    Hand-tuned per baseline and dataset; directly scales the added losses in Eq. (7), so it materially affects the reported FID/IS.
  • τ (temperature in SupCon loss) = not reported
    The temperature in Eq. (1) is a free hyperparameter; its value is never given, yet it controls the sharpness of the contrastive distribution.
assumptions (3)
  • domain assumption Class labels are accurate and available for all training images and are a sufficient proxy for semantic similarity between images and between image-text pairs.
    Invoked in Sec. 3.1 sampling strategy and loss definitions; if labels are noisy or too coarse, positive pairs become semantically unrelated and the contrastive loss would hurt generation.
  • domain assumption The frozen pre-trained encoders provide meaningful representations of generated (fake) images, so gradients from the supervised contrastive loss on those representations improve the generator.
    Invoked in Sec. 3.2; the encoders are inference-only, so the only path for L_sup to affect the generator is through these fixed features. If those features are not sensitive to semantic differences in fake images, the loss provides little gradient signal.
  • standard math Standard GAN training assumptions (adversarial convergence, optimization stability) hold in the baseline methods.
    The paper builds on AttnGAN, DM-GAN, SSA-GAN, GALIP without modifying their core GAN losses; it assumes those training procedures work as published.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Framework For Image Synthesis Using Supervised Contrastive Learning." pith.science (2026). https://pith.science/paper/W6UIF5HQ

@misc{pith2026241203957,
  author       = {Pith},
  title        = {Pith review of: A Framework For Image Synthesis Using Supervised Contrastive Learning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/W6UIF5HQ}},
  note         = {Machine review of arXiv:2412.03957}
}
read the original abstract

Text-to-image (T2I) generation aims at producing realistic images corresponding to text descriptions. Generative Adversarial Network (GAN) has proven to be successful in this task. Typical T2I GANs are 2 phase methods that first pretrain an inter-modal representation from aligned image-text pairs and then use GAN to train image generator on that basis. However, such representation ignores the inner-modal semantic correspondence, e.g. the images with same label. The semantic label in priory describes the inherent distribution pattern with underlying cross-image relationships, which is supplement to the text description for understanding the full characteristics of image. In this paper, we propose a framework leveraging both inter- and inner-modal correspondence by label guided supervised contrastive learning. We extend the T2I GANs to two parameter-sharing contrast branches in both pretraining and generation phases. This integration effectively clusters the semantically similar image-text pair representations, thereby fostering the generation of higher-quality images. We demonstrate our framework on four novel T2I GANs by both single-object dataset CUB and multi-object dataset COCO, achieving significant improvements in the Inception Score (IS) and Frechet Inception Distance (FID) metrics of imagegeneration evaluation. Notably, on more complex multi-object COCO, our framework improves FID by 30.1%, 27.3%, 16.2% and 17.1% for AttnGAN, DM-GAN, SSA-GAN and GALIP, respectively. We also validate our superiority by comparing with other label guided T2I GANs. The results affirm the effectiveness and competitiveness of our approach in advancing the state-of-the-art GAN for T2I generation

Figures

Figures reproduced from arXiv: 2412.03957 by the authors.

Figure 1
Figure 1. Pre-training phase. Our data sampling strategy initiates two contrast branches with shared parameters to separately encode the image-text pairs of same label. The original Loss is consistent to the method our framework applied on. The supervised con￾trastive loss works on quadruple of image and text representations from both branches. 3.1 Supervised Contrastive Learning for Pre-training Typical T2I GANs pre-train th… view at source ↗
Figure 2
Figure 2. GAN training phase. Same as pre-training phase, we use two parameter-sharing T2I GAN branches to contrast the text-image pairs sharing same label. The supervised contrastive loss is performed on quadruple of text and generated fake image represen￾tations from two branches. In this phase, the pre-trained encoders are inference-only. The final objective function for the pre-training phase is a co-op of origin loss and… view at source ↗
Figure 3
Figure 3. Qualitative comparison on CUB and COCO datasets for DM-GAN and SSA￾GAN baselines w/o the utilization of our framework (denoted as ”+SCL”). The input text descriptions are given in the first row and the corresponding generated images from different methods are shown in the same column. The left 4 columns are from CUB, and right 4 columns from COCO. Implementation Details We apply our framework to four novel baselines… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 15 canonical work pages

  1. [1]

    In: International conference on machine learning

    Chen, T., Kornblith, S., Norouzi, M., Hinton, G.: A simple framework for con- trastive learning of visual representations. In: International conference on machine learning. pp. 1597–1607. PMLR (2020)

    2020

  2. [2]

    In: 2009 IEEE Computer Society Con- ference on Computer Vision and Pattern Recognition (CVPR 2009), 20- 25 June 2009, Miami, Florida, USA

    Deng, J., Dong, W., Socher, R., Li, L., Li, K., Fei-Fei, L.: Imagenet: A large-scale hierarchical image database. In: 2009 IEEE Computer Society Con- ference on Computer Vision and Pattern Recognition (CVPR 2009), 20- 25 June 2009, Miami, Florida, USA. pp. 248–255. IEEE Computer Soci- ety (2009). https://doi.org/10.1109/CVPR.2009.5206848, https://doi.org/...

    arXiv 2009

  3. [3]

    In: Ghahramani, Z., Welling, M., Cortes, C., Lawrence, N.D., Weinberger, K.Q

    Goodfellow, I.J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., Courville, A.C., Bengio, Y.: Generative adversarial nets. In: Ghahramani, Z., Welling, M., Cortes, C., Lawrence, N.D., Weinberger, K.Q. (eds.) Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, December 8...

    2014

  4. [4]

    ´A., Guo, Z., Azar, M.G., Piot, B., Kavukcuoglu, K., Munos, R., Valko, M.: Bootstrap your own latent - A new approach to self-supervised learning

    Grill, J., Strub, F., Altch´ e, F., Tallec, C., Richemond, P.H., Buchatskaya, E., Do- ersch, C., Pires, B. ´A., Guo, Z., Azar, M.G., Piot, B., Kavukcuoglu, K., Munos, R., Valko, M.: Bootstrap your own latent - A new approach to self-supervised learning. In: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M., Lin, H. (eds.) Advances in Neural Information...

    work page 2020

  5. [5]

    In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2020, Seattle, W A, USA, June 13-19, 2020

    He, K., Fan, H., Wu, Y., Xie, S., Girshick, R.B.: Momentum contrast for unsupervised visual representation learning. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2020, Seattle, W A, USA, June 13-19, 2020. pp. 9726–9735. Computer Vision Foundation / IEEE (2020). https://doi.org/10.1109/CVPR42600.2020.00975, https://doi.org/...

    arXiv 2020

  6. [6]

    Advances in neural information processing systems 33, 18661–18673 (2020)

    Khosla, P., Teterwak, P., Wang, C., Sarna, A., Tian, Y., Isola, P., Maschinot, A., Liu, C., Krishnan, D.: Supervised contrastive learning. Advances in neural information processing systems 33, 18661–18673 (2020)

    2020

  7. [7]

    arXiv preprint arXiv:2304.02643 (2023)

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.Y., et al.: Segment anything. arXiv preprint arXiv:2304.02643 (2023)

    arXiv 2023

  8. [8]

    In: IEEE Confer- ence on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019

    Li, W., Zhang, P., Zhang, L., Huang, Q., He, X., Lyu, S., Gao, J.: Object- driven text-to-image synthesis via adversarial training. In: IEEE Confer- ence on Computer Vision and Pattern Recognition, CVPR 2019, Long Beach, CA, USA, June 16-20, 2019. pp. 12174–12182. Computer Vision Foundation / IEEE (2019). https://doi.org/10.1109/CVPR.2019.01245, http://op...

    work page arXiv 2019

Show all 30 references
  1. [10]

    CoRR abs/1705.02894 (2017), http:// arxiv.org/abs/1705.02894

    Lim, J.H., Ye, J.C.: Geometric GAN. CoRR abs/1705.02894 (2017), http:// arxiv.org/abs/1705.02894

    2017 arXiv

  2. [11]

    In: Computer Vision– ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13

    Lin, T.Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Doll´ ar, P., Zitnick, C.L.: Microsoft coco: Common objects in context. In: Computer Vision– ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V 13. pp. 740–7...

    2014

  3. [12]

    IEEE Transactions on Neural Networks and Learning Systems (2022)

    Ma lki´ nski, M., Ma´ ndziuk, J.: Multi-label contrastive learning for abstract visual reasoning. IEEE Transactions on Neural Networks and Learning Systems (2022)

    2022

  4. [13]

    CoRR abs/1807.03748 (2018), http://arxiv.org/abs/1807

    van den Oord, A., Li, Y., Vinyals, O.: Representation learning with contrastive predictive coding. CoRR abs/1807.03748 (2018), http://arxiv.org/abs/1807. 03748

    2018 arXiv

  5. [14]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Qiao, T., Zhang, J., Xu, D., Tao, D.: Mirrorgan: Learning text-to-image genera- tion by redescription. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 1505–1514 (2019)

    2019

  6. [15]

    In: Meila, M., Zhang, T

    Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., Krueger, G., Sutskever, I.: Learning transferable visual models from natural language supervision. In: Meila, M., Zhang, T. (eds.) Proceedings of the 38th Int...

    2021

  7. [16]

    In: International conference on machine learning

    Reed, S., Akata, Z., Yan, X., Logeswaran, L., Schiele, B., Lee, H.: Generative ad- versarial text to image synthesis. In: International conference on machine learning. pp. 1060–1069. PMLR (2016)

    2016

  8. [17]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Ruan, S., Zhang, Y., Zhang, K., Fan, Y., Tang, F., Liu, Q., Chen, E.: Dae-gan: Dynamic aspect-aware gan for text-to-image synthesis. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 13960–13969 (2021)

    2021

  9. [18]

    Advances in neural information processing systems 29 (2016)

    Salimans, T., Goodfellow, I., Zaremba, W., Cheung, V., Radford, A., Chen, X.: Improved techniques for training gans. Advances in neural information processing systems 29 (2016)

    2016

  10. [19]

    IEEE Trans

    Schuster, M., Paliwal, K.K.: Bidirectional recurrent neural networks. IEEE Trans. Signal Process. 45(11), 2673–2681 (1997). https://doi.org/10.1109/78.650093, https://doi.org/10.1109/78.650093 A Framework For Image Synthesis Using Supervised Contrastive Learning 15

    1997 doi

  11. [20]

    In: 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016

    Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., Wojna, Z.: Rethinking the inception architecture for computer vision. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016, Las Vegas, NV, USA, June 27-30, 2016. pp. 2818–2826. IEEE Computer Society (...

    2016 doi

  12. [21]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24, 2023

    Tao, M., Bao, B., Tang, H., Xu, C.: GALIP: generative adversarial clips for text- to-image synthesis. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2023, Vancouver, BC, Canada, June 17-24, 2023. pp. 14214– 14223. IEEE (2023). https://doi.org/10.1109/...

    2023

  13. [23]

    arXiv preprint arXiv:1708.08819 (2017)

    Unterthiner, T., Nessler, B., Seward, C., Klambauer, G., Heusel, M., Ramsauer, H., Hochreiter, S.: Coulomb gans: Provably optimal nash equilibria via potential fields. arXiv preprint arXiv:1708.08819 (2017)

    2017 arXiv

  14. [24]

    california institute of technology (2011)

    Wah, C., Branson, S., Welinder, P., Perona, P., Belongie, S.: The caltech-ucsd birds-200-2011 dataset. california institute of technology (2011)

    2011

  15. [25]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Xu, T., Zhang, P., Huang, Q., Zhang, H., Gan, Z., Huang, X., He, X.: Attngan: Fine-grained text to image generation with attentional generative adversarial net- works. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 1316–1324 (2018)

    2018

  16. [26]

    In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022

    Yang, J., Li, C., Zhang, P., Xiao, B., Liu, C., Yuan, L., Gao, J.: Unified contrastive learning in image-text-label space. In: IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022, New Orleans, LA, USA, June 18-24, 2022. pp. 19141–19151. IEEE (2022). https...

    2022

  17. [27]

    arXiv preprint arXiv:2107.02423 (2021)

    Ye, H., Yang, X., Takac, M., Sunderraman, R., Ji, S.: Improving text-to-image synthesis using contrastive learning. arXiv preprint arXiv:2107.02423 (2021)

    2021 arXiv

  18. [28]

    In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition

    Yin, G., Liu, B., Sheng, L., Yu, N., Wang, X., Shao, J.: Semantics disentangling for text-to-image generation. In: Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition. pp. 2327–2336 (2019)

    2019

  19. [29]

    In: Meila, M., Zhang, T

    Zbontar, J., Jing, L., Misra, I., LeCun, Y., Deny, S.: Barlow twins: Self-supervised learning via redundancy reduction. In: Meila, M., Zhang, T. (eds.) Proceedings of the 38th International Conference on Machine Learning, ICML 2021, 18-24 July 2021, Virtual Event. Proceedings ...

    2021

  20. [30]

    In: Proceedings of the IEEE international conference on computer vi- sion

    Zhang, H., Xu, T., Li, H., Zhang, S., Wang, X., Huang, X., Metaxas, D.N.: Stack- gan: Text to photo-realistic image synthesis with stacked generative adversarial networks. In: Proceedings of the IEEE international conference on computer vi- sion. pp. 5907–5915 (2017)

    2017

  21. [31]

    IEEE transactions on pattern analysis and machine intelligence 41(8), 1947–1962 (2018) 16 Yibin Liu ∗, Jianyu Zhang ∗, Li Zhang, Shijian Li †, and Gang Pan

    Zhang, H., Xu, T., Li, H., Zhang, S., Wang, X., Huang, X., Metaxas, D.N.: Stack- gan++: Realistic image synthesis with stacked generative adversarial networks. IEEE transactions on pattern analysis and machine intelligence 41(8), 1947–1962 (2018) 16 Yibin Liu ∗, Jianyu Zhang ∗...

    2018

  22. [32]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Zhu, M., Pan, P., Chen, W., Yang, Y.: Dm-gan: Dynamic memory generative ad- versarial networks for text-to-image synthesis. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 5802–5810 (2019)

    2019

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.