Pith. sign in

REVIEW 3 major objections 5 minor 48 references

Photovoltaic Defect Image Generator with Boundary Alignment Smoothing Constraint for Domain Shift Mitigation

T0 review · 3 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A Stable Diffusion-based generator for photovoltaic defect images lowers FID by 19.16 over StyleGAN3 and lifts YOLOv5 mean average precision by 5.8–6.3 points on held-out production lines.

desk verdict A sensible recombination of known diffusion tricks for PV defect generation, with a genuine held-out evaluation, but the domain-shift claim needs a data-matched control and some mandatory baselines. read the letter →

arxiv 2505.06117 v1 pith:JZVTNRZF submitted 2025-05-09 cs.CV

Dongying Li , Binyi Su , Hua Zhang , Yong Li , Haiyong Chen This is my paper · ORCID
classification cs.CV
keywords photovoltaicdefectdetectionelectroluminescenceimagingStableDiffusiontext-to-imagegenerationdomainshiftdataaugmentationobjectendogenous
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 proposes PDIG, a photovoltaic defect image generator built on a frozen Stable Diffusion model, and argues that its synthetic electroluminescence defect images are realistic and diverse enough to reduce the endogenous domain shift between production lines — the distribution change caused by differing resolution, brightness, grid-line spacing, and defect morphology. PDIG learns a text embedding for each defect concept, injects real defect-image patches through a cross-disentangled attention adaptor, and at inference steers generation inside user-specified boxes with boundary-smoothing constraints. The authors report an image-fidelity score (FID) of 20.90, beating the second-best StyleGAN3 baseline by 19.16 points, and show that augmenting a single production line's training set with PDIG images raises YOLOv5 mean average precision (mAP) by 5.8 and 6.3 points on two held-out line groups. For manufacturers, this would provide a data-augmentation route to defect detectors that work across production lines without recollecting and re-labeling images from each line.

What carries the argument

The central object is the cross-disentangled attention in the Lightweight Industrial Style Adaptor (LISA). Caption tokens supply the query while image tokens — produced by cropping defect regions, encoding them with a pretrained CLIP image encoder, and projecting them through a linear layer — supply key and value, with the two attention branches combined as $$\bar{H}=\$\alpha$\,\mathrm{Attn}(Q_p,K_z,V_z)+\$\beta$\,\mathrm{Attn}(Q_p,K_p,V_p).$$ This is what lets a few industrial defect images steer the frozen diffusion model without retraining it. The second load-bearing mechanism is the Text-Image Dual-Space Constraints (TIDSC) module, whose total loss $$L=L_{In}+L_{Out}+L_{BASC}$$ combines masked internal and external attention constraints with axis-projected boundary matching and first- and second-order smoothing of the attention map; this enforces box-localized generation and softer defect boundaries.

What would settle it

An independent replication of the YOLOv5 experiment on the same ELES split that cannot reproduce the reported +5.8 and +6.3 mAP gains on EL group2 and EL group3 would falsify the central claim that PDIG-augmented data mitigates endogenous shift.

Watch

Extended reading notes

Core claim

The paper asserts that PDIG — a two-stage pipeline built on a frozen Stable Diffusion model — can generate photovoltaic electroluminescence defect images that are realistic enough and diverse enough to reduce endogenous domain shift in defect detection. During training, the Semantic Concept Embedding (SCE) module learns a text token for PV EL imaging and each defect type, while the Lightweight Industrial Style Adaptor (LISA) crops defect patches, encodes them with a pretrained CLIP image encoder, and injects them as image tokens through cross-disentangled attention parallel to text tokens. At inference, the Text-Image Dual-Space Constraints (TIDSC) module steers cross-attention maps inside user-specified boxes, adding axis-projected boundary alignment and smoothing terms plus internal and external mask constraints, so the generated defect sits at the requested location. The authors report average FID of 20.90 versus 40.06 for StyleGAN3, a 19.16-point gap, and show that adding 200 generated images per defect type to the EL group1 training set raises YOLOv5 mean average precision by 5.8 and 6.3 points on EL group2 and EL group3.

Load-bearing premise

The load-bearing premise is that the pretrained CLIP image encoder, trained on natural images, can represent the fine-grained morphology of PV EL defects (cracks, broken gates, black spots) well enough that the image tokens injected through LISA actually guide diffusion; if those features are not captured, LISA cannot improve generation.

Editorial extensions

If this is right

  • A detector trained only on EL group1 images can, after augmentation with PDIG output, detect defects on later production lines: YOLOv5 mAP rises from 57.6 to 63.4 on EL group2 and from 42.2 to 48.5 on EL group3.
  • YOLOv8 also benefits, gaining 1.8 mAP points on both EL group2 and EL group3, so the improvement is not tied to a single detector family.
  • The TIDSC constraint localizes generated defects with an average 84.74% pass rate and 8.37% box spillover, so generated images arrive with usable annotation priors.
  • Ablation results attribute the FID gain mainly to LISA: LISA alone lowers FID from 197.03 to 48.64, adding SCE brings it to 36.78, and the full pipeline reaches 20.90.

Reading between the lines

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

  • A likely transferable corollary of this design is that the frozen-Stable-Diffusion plus cross-disentangled image-adapter recipe can be moved to other small-target industrial defect domains (welds, circuits, textiles), since nothing in LISA is PV-specific.
  • Because LISA's image tokens come from a CLIP encoder trained on natural images, fine crack morphology may be underrepresented; swapping in a domain-specific encoder while keeping everything else fixed is a direct way to test whether the reported gains have headroom.
  • The paper demonstrates gains relative to training on a single line; a stronger implicit claim is that generated data approaches the value of real target-line data, which could be tested by comparing PDIG-augmented group1 training against adding a small real sample from group2 or group3.
  • Since box spillover remains at 8.37%, weighting the boundary-smoothing terms more heavily or adding post-hoc mask erosion could be tried to reduce overflow without retraining the base model.
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 manuscript proposes PDIG, a Stable-Diffusion-based generator for photovoltaic electroluminescence (EL) defect images, with three main components: a Semantic Concept Embedding (SCE) module that learns text embeddings for industrial defect concepts, a Lightweight Industrial Style Adaptor (LISA) that injects image features through cross-disentangled attention, and a Text-Image Dual-Space Constraints (TIDSC) module that constrains cross-attention maps at inference so that defects are generated at user-specified bounding boxes. The model is trained only on EL group1 of the ELES dataset. The authors evaluate generation quality with FID/IS against DCGAN, StyleGAN3, Textual Inversion, IP-Adapter, and T2I-Adapter, and they evaluate downstream defect detection by adding 1,000 generated, automatically annotated images to the YOLOv5/YOLOv8 training set. The headline results are an average FID 19.16 lower than the second-best method and YOLOv5 mAP gains of 5.8 and 6.3 points on EL group2 and EL group3, respectively.

Significance. PDIG addresses a real industrial need: generating location-annotated PV defect images under limited training data. The architecture is reasonable and well motivated, and the idea of combining a frozen Stable Diffusion backbone with a lightweight image adapter plus a training-free inference constraint is practically attractive because it produces annotations for downstream detection. If the reported gains survive controlled evaluation, the method would be a useful contribution to industrial defect data augmentation and domain-shift mitigation. The paper is weakened, however, by the absence of code, error bars or significance tests, a data-volume-matched control for the detection experiment, and direct comparisons with two closely related methods that are cited in the text. The central claim of effective domain-shift mitigation is therefore plausible but not yet established by the evidence as presented.

major comments (3)
  1. [Section IV-F, Table V] The central claim that PDIG "effectively mitigates endogenous shifts" is underdetermined because the only comparison is 6,652 real EL group1 images versus 6,652 real plus 1,000 generated images. Adding any additional diverse data, including repeated real images, standard geometric/color augmentations, or images from StyleGAN3, IP-Adapter, or T2I-Adapter as evaluated in Table II, can improve mAP on shifted test sets, so the observed +5.8/+6.3 does not isolate the proposed domain-shift mechanism. Please add data-volume-matched controls with real images, standard augmentations, and images from the other generators, and report per-seed error bars or significance tests. The per-class entries are also not uniformly positive (e.g., YOLOv5 broken_gate on EL group1 drops by 3.3 and YOLOv8 black_spot on EL group2 drops by 1.4), which further indicates that the average gains need statistical support.
  2. [Section II-C and Section IV-D] AnomalyDiffusion [35] and BoxDiff [44] are cited as the direct motivation for the TIDSC module, but neither method appears in the experimental comparison. BoxDiff is a training-free box-constrained diffusion method and AnomalyDiffusion is a few-shot anomaly image generation method with spatial embedding, so both are natural baselines for the position-conditioned generation claim reported in Table IV and for the generation-quality comparison in Table II. Without these comparisons, the paper does not support the statement that the proposed spatial-control mechanism improves over existing spatial-control mechanisms.
  3. [Section IV-D] The reference distribution used to compute FID is never stated. If the FID in Table II is computed against EL group1 real images, a lower FID indicates better fidelity to the source domain, which is not the same as better generalization to the shifted EL group2 and EL group3 test distributions. The paper should state the reference set explicitly and, ideally, report FID against all three EL groups so that the generation-quality metric is informative about the claimed domain-shift mitigation.
minor comments (5)
  1. [Section IV-E3, Table III] The text says that combining SCE with TIDSC decreases IS by 0.23 relative to TIDSC alone, but Table III shows a change from 3.72 to 2.05, a decrease of 1.67; please correct this inconsistency.
  2. [Eq. (19)] The FID formula is malformed as printed, with mismatched parentheses and undefined covariance matrices; as written it cannot be verified. Please provide the standard FID definition with explicit covariance notation.
  3. [Section II-C and Section III-C] The acronym TIDSC is expanded inconsistently: Section III-C uses "Text-Image Dual-Space Constraints" while Section II-C uses "Targeted Instance-Dependent Spatial Consistency"; please use one expansion throughout.
  4. [Section III-A, Eq. (2)] In the SCE objective, the marker embedding vector v is introduced but the equation does not show how v enters the text encoder mapping τθ(y) or the prompt. Please define the vocabulary insertion mechanism explicitly.
  5. [Section III-C, Eqs. (11) and (14)] The term "tokp" appears to be a typo for the sampling/top-k operator used in the other constraints; please correct the notation.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the FID and downstream detection claims are tested against external benchmarks and held-out production-line data; the only overlapping-author citation is the dataset source [5], which is not load-bearing.

full rationale

The paper's central claims are empirical: (1) PDIG achieves a lower FID than other generators, and (2) adding PDIG-generated images to the EL group1 training set improves YOLOv5/YOLOv8 mAP on held-out EL group2 and EL group3 production lines. Neither claim is derived from the method's own equations in a way that makes the outcome true by construction. The SCE optimization in Eq. (2) is the standard textual-inversion objective, LISA in Eqs. (3)-(4) is a cross-attention injection that concatenates image and text tokens, and TIDSC in Eqs. (5)-(18) is an inference-time loss on cross-attention maps; none of these define the reported FID or detection mAP as an input. The evaluation uses external pretrained models (Stable Diffusion, CLIP) and external metrics (FID, IS, mAP), and the detection test sets are from different production lines not used for training. The only author-overlapping citation is the ELES dataset source [5], which provides the benchmark split; the paper also independently visualizes the domain shift with t-SNE and density plots in Fig. 1, so the citation is not the sole justification. The absence of a data-volume-matched control with real images or other generators in Table V weakens the causal attribution to domain-shift mitigation, but that is an experimental-design limitation, not a circular reduction. Accordingly, no circular step is identified; the minor self-citation to the dataset source is not load-bearing.

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

No new physical or conceptual entities are introduced; SCE, LISA, and TIDSC are modules, not invented entities in the sense of new forces or particles. The free parameters are the unspecified or contradictory hyperparameters in the loss and attention weighting.

free parameters (4)
  • alpha, beta cross-attention weights = alpha=0, beta=1.0 in image-prompt-only mode (contradicts Eq. 4 semantics)
    Chosen by hand in implementation details; the stated values would disable the image branch, which is inconsistent with the module's purpose.
  • lambda1, lambda2 smoothing weights = not specified
    Weights for first-order and second-order smoothing terms in Eq. (17); values are never given, making the loss scale ambiguous.
  • top-k count P = not specified
    Used in internal/external attention constraints (Eqs. 5-8); number of selected top elements not provided.
  • sampling count L = not specified
    Number of uniform samples in boundary alignment losses (Eqs. 11, 14); not specified.
assumptions (4)
  • domain assumption Stable Diffusion (SDv1.5) pretrained on LAION-5B provides a useful prior for generating PV EL defect images.
    Section III-A uses SD as backbone, and Fig. 2 shows naive SD fails on this domain, so the prior is necessary but not sufficient.
  • domain assumption CLIP ViT-H/14 image and text encoders can capture PV EL defect semantics relevant to generation.
    LISA uses CLIP image encoder; SCE uses CLIP text encoder. The domain gap between natural images and EL images is large and not verified.
  • domain assumption Cross-attention maps in the frozen U-Net correlate with object position, so optimizing the latent against attention masks (TIDSC) can place generated defects in specified boxes.
    This is an empirical property used in BoxDiff [44], adopted here without verification for PV tokens.
  • domain assumption FID, IS, and YOLO mAP are appropriate metrics for generation quality and detection improvement.
    Standard metrics, but FID with only 1000 generated images is high-variance and IS is known to be unreliable for small defect objects.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Photovoltaic Defect Image Generator with Boundary Alignment Smoothing Constraint for Domain Shift Mitigation." pith.science (2026). https://pith.science/paper/JZVTNRZF

@misc{pith2026250506117,
  author       = {Pith},
  title        = {Pith review of: Photovoltaic Defect Image Generator with Boundary Alignment Smoothing Constraint for Domain Shift Mitigation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JZVTNRZF}},
  note         = {Machine review of arXiv:2505.06117}
}
read the original abstract

Accurate defect detection of photovoltaic (PV) cells is critical for ensuring quality and efficiency in intelligent PV manufacturing systems. However, the scarcity of rich defect data poses substantial challenges for effective model training. While existing methods have explored generative models to augment datasets, they often suffer from instability, limited diversity, and domain shifts. To address these issues, we propose PDIG, a Photovoltaic Defect Image Generator based on Stable Diffusion (SD). PDIG leverages the strong priors learned from large-scale datasets to enhance generation quality under limited data. Specifically, we introduce a Semantic Concept Embedding (SCE) module that incorporates text-conditioned priors to capture the relational concepts between defect types and their appearances. To further enrich the domain distribution, we design a Lightweight Industrial Style Adaptor (LISA), which injects industrial defect characteristics into the SD model through cross-disentangled attention. At inference, we propose a Text-Image Dual-Space Constraints (TIDSC) module, enforcing the quality of generated images via positional consistency and spatial smoothing alignment. Extensive experiments demonstrate that PDIG achieves superior realism and diversity compared to state-of-the-art methods. Specifically, our approach improves Frechet Inception Distance (FID) by 19.16 points over the second-best method and significantly enhances the performance of downstream defect detection tasks.

Figures

Figures reproduced from arXiv: 2505.06117 by the authors.

Figure 1
Figure 1. The published ELES dataset exhibits issues of domain shift and [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The images generated by the SD V1.5 model with the prompt “PV [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The PDIG employs a two-stage generation approach. During the training stage, the Semantic Concept Embedding (SCE) module and the Industrial [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Illustration of cross-disentangled attention between Text and Image. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Distribution of defect categories in each production line of the dataset. [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Qualitative comparison of ours with other models. Generated target defects are highlighted in red boxes. [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Comparison of t-SNE Dimensionality Reduction and Probability [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 9
Figure 9. Figure 9: Visualization Results of the TIDSC Module. For different selected [PITH_FULL_IMAGE:figures/full_fig_p009_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 27 canonical work pages

  1. [35]

    Anomalydiffusion: Few-shot anomaly image generation with diffusion model,

    T. Hu, J. Zhang, R. Yi, Y . Du, X. Chen, L. Liu, Y . Wang, and C. Wang, “Anomalydiffusion: Few-shot anomaly image generation with diffusion model,” in Proceedings of the AAAI conference on artificial intelligence , vol. 38, no. 8, 2024, pp. 8526–8534

    work page 2024

  2. [44]

    Boxdiff: Text-to-image synthesis with training-free box-constrained diffusion,

    J. Xie, Y . Li, Y . Huang, H. Liu, W. Zhang, Y . Zheng, and M. Z. Shou, “Boxdiff: Text-to-image synthesis with training-free box-constrained diffusion,” in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 7452–7461

    work page 2023

  3. [1]

    Investi- gating the impact of cracks on solar cells performance: Analysis based on nonuniform and uniform crack dis- tributions,

    M. Dhimish, V . d’Alessandro, and S. Daliento, “Investi- gating the impact of cracks on solar cells performance: Analysis based on nonuniform and uniform crack dis- tributions,” IEEE Transactions on Industrial Informatics, vol. 18, no. 3, pp. 1684–1693, 2021

    work page 2021

  4. [2]

    Deep learning-based solar-cell manufacturing defect de- tection with complementary attention network,

    B. Su, H. Chen, P. Chen, G. Bian, K. Liu, and W. Liu, “Deep learning-based solar-cell manufacturing defect de- tection with complementary attention network,” IEEE Transactions on Industrial informatics , vol. 17, no. 6, pp. 4084–4095, 2020

    work page 2020

  5. [3]

    Baf-detector: An efficient cnn-based detector for photovoltaic cell defect detection,

    B. Su, H. Chen, and Z. Zhou, “Baf-detector: An efficient cnn-based detector for photovoltaic cell defect detection,” IEEE Transactions on Industrial Electronics , vol. 69, no. 3, pp. 3161–3171, 2021

    work page 2021

  6. [4]

    Sncf-net: Scale-aware neighborhood correlation feature network for hotspot defect detection of photovoltaic farms,

    S. Zhao, H. Chen, C. Wang, and S. Shi, “Sncf-net: Scale-aware neighborhood correlation feature network for hotspot defect detection of photovoltaic farms,” Mea- surement, vol. 206, p. 112342, 2023

    work page 2023

  7. [5]

    Ssn: Shift suppression network for endogenous shift of photovoltaic defect detection,

    S. Zhao, H. Chen, C. Wang, and Z. Zhang, “Ssn: Shift suppression network for endogenous shift of photovoltaic defect detection,” IEEE Transactions on Industrial Infor- matics, 2023

    work page 2023

  8. [6]

    Unveiling the potential of progressive training diffusion model for defect image generation and recogni- tion in industrial processes,

    Y . Wang, Z. Zhou, X. Tan, Y . Pan, J. Yuan, Z. Qiu, and C. Liu, “Unveiling the potential of progressive training diffusion model for defect image generation and recogni- tion in industrial processes,” Neurocomputing, vol. 592, p. 127837, 2024

    work page 2024

Show all 48 references
  1. [7]

    A survey on image data augmentation for deep learning,

    C. Shorten and T. M. Khoshgoftaar, “A survey on image data augmentation for deep learning,” Journal of big data, vol. 6, no. 1, pp. 1–48, 2019

    2019

  2. [8]

    Generative adversarial networks: An overview,

    A. Creswell, T. White, V . Dumoulin, K. Arulkumaran, B. Sengupta, and A. A. Bharath, “Generative adversarial networks: An overview,” IEEE signal processing maga- zine, vol. 35, no. 1, pp. 53–65, 2018

    2018

  3. [9]

    Defect- gan: High-fidelity defect synthesis for automated defect inspection,

    G. Zhang, K. Cui, T.-Y . Hung, and S. Lu, “Defect- gan: High-fidelity defect synthesis for automated defect inspection,” in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision , 2021, pp. 2524–2534

    2021

  4. [10]

    Defect sample image generation method based on gans in diamond tool defect detection,

    C. Zhao, W. Xue, W.-P. Fu, Z.-Q. Li, and X. Fang, “Defect sample image generation method based on gans in diamond tool defect detection,” IEEE Transactions on Instrumentation and Measurement , vol. 72, pp. 1–9, 2023

    2023

  5. [11]

    A survey on generative diffusion models,

    H. Cao, C. Tan, Z. Gao, Y . Xu, G. Chen, P.-A. Heng, and S. Z. Li, “A survey on generative diffusion models,” IEEE Transactions on Knowledge and Data Engineering, 2024

    2024

  6. [12]

    Imagpose: A unified conditional framework for pose-guided person generation,

    F. Shen and J. Tang, “Imagpose: A unified conditional framework for pose-guided person generation,” Advances in neural information processing systems , vol. 37, pp. 6246–6266, 2024

    2024

  7. [13]

    Imagdressing-v1: Customizable virtual dressing,

    F. Shen, X. Jiang, X. He, H. Ye, C. Wang, X. Du, Z. Li, and J. Tang, “Imagdressing-v1: Customizable virtual dressing,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 39, no. 7, 2025, pp. 6795– 6804

    2025

  8. [14]

    High-resolution image synthesis with la- tent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with la- tent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2022, pp. 10 684–10 695

    2022

  9. [15]

    Sgdm: An adaptive style-guided diffusion model for personalized text to image generation,

    Y . Xu, X. Xu, H. Gao, and F. Xiao, “Sgdm: An adaptive style-guided diffusion model for personalized text to image generation,” IEEE Transactions on Multimedia , vol. 26, pp. 9804–9813, 2024

    2024

  10. [16]

    Adding con- ditional control to text-to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding con- ditional control to text-to-image diffusion models,” in Proceedings of the IEEE/CVF international conference on computer vision , 2023, pp. 3836–3847

    2023

  11. [17]

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

    H. Ye, J. Zhang, S. Liu, X. Han, and W. Yang, “Ip-adapter: Text compatible image prompt adapter for text-to-image diffusion models,” arXiv preprint arXiv:2308.06721, 2023

    2023 arXiv

  12. [18]

    T2i-adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models,

    C. Mou, X. Wang, L. Xie, Y . Wu, J. Zhang, Z. Qi, and Y . Shan, “T2i-adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models,” in Proceedings of the AAAI conference on artificial intelligence, vol. 38, no. 5, 2024, pp. 4296–4304

    2024

  13. [19]

    Imaggarment-1: Fine-grained garment gener- ation for controllable fashion design,

    F. Shen, J. Yu, C. Wang, X. Jiang, X. Du, and J. Tang, “Imaggarment-1: Fine-grained garment gener- ation for controllable fashion design,” arXiv preprint JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 11 arXiv:2504.13176, 2025

    2021 arXiv

  14. [20]

    Long-term talkingface generation via motion-prior conditional diffusion model,

    F. Shen, C. Wang, J. Gao, Q. Guo, J. Dang, J. Tang, and T.-S. Chua, “Long-term talkingface generation via motion-prior conditional diffusion model,” arXiv preprint arXiv:2502.09533, 2025

    2025 arXiv

  15. [21]

    Region-and strength-controllable gan for defect generation and seg- mentation in industrial images,

    S. Niu, B. Li, X. Wang, and Y . Peng, “Region-and strength-controllable gan for defect generation and seg- mentation in industrial images,” IEEE Transactions on Industrial Informatics , vol. 18, no. 7, pp. 4531–4541, 2021

    2021

  16. [22]

    Uncertainty-aware unsupervised domain adaptation in object detection,

    D. Guan, J. Huang, A. Xiao, S. Lu, and Y . Cao, “Uncertainty-aware unsupervised domain adaptation in object detection,” IEEE Transactions on Multimedia , vol. 24, pp. 2502–2514, 2022

    2022

  17. [23]

    Dp-gan: A transmission line bolt defects generation network based on dual discriminator architecture and pseudo-enhancement strategy,

    K. Zhang, Y . Xiao, J. Wang, M. Du, X. Guo, R. Zhou, C. Shi, and Z. Zhao, “Dp-gan: A transmission line bolt defects generation network based on dual discriminator architecture and pseudo-enhancement strategy,” IEEE Transactions on Power Delivery , 2024

    2024

  18. [24]

    Semantic-supervised infrared and visible image fusion via a dual-discriminator generative adversarial network,

    H. Zhou, W. Wu, Y . Zhang, J. Ma, and H. Ling, “Semantic-supervised infrared and visible image fusion via a dual-discriminator generative adversarial network,” IEEE Transactions on Multimedia, vol. 25, pp. 635–648, 2023

    2023

  19. [25]

    Adabalgan: An improved generative adversarial network with imbalanced learning for wafer defective pattern recognition,

    J. Wang, Z. Yang, J. Zhang, Q. Zhang, and W.-T. K. Chien, “Adabalgan: An improved generative adversarial network with imbalanced learning for wafer defective pattern recognition,” IEEE Transactions on Semiconduc- tor Manufacturing, vol. 32, no. 3, pp. 310–319, 2019

    2019

  20. [26]

    Conditional generative ad- versarial nets,

    M. Mirza and S. Osindero, “Conditional generative ad- versarial nets,” arXiv preprint arXiv:1411.1784 , 2014

    2014 arXiv

  21. [27]

    Detection of apple lesions in orchards based on deep learning methods of cyclegan and yolov3-dense,

    Y . Tian, G. Yang, Z. Wang, E. Li, and Z. Liang, “Detection of apple lesions in orchards based on deep learning methods of cyclegan and yolov3-dense,”Journal of Sensors, vol. 2019, no. 1, p. 7630926, 2019

    2019

  22. [28]

    Defect image sample generation with gan for improving defect recog- nition,

    S. Niu, B. Li, X. Wang, and H. Lin, “Defect image sample generation with gan for improving defect recog- nition,” IEEE Transactions on Automation Science and Engineering, vol. 17, no. 3, pp. 1611–1622, 2020

    2020

  23. [29]

    Dg2gan: improving defect recognition performance with generated defect image sample,

    F. Deng, J. Luo, L. Fu, Y . Huang, J. Chen, N. Li, J. Zhong, and T. L. Lam, “Dg2gan: improving defect recognition performance with generated defect image sample,” Scientific Reports , vol. 14, no. 1, p. 14787, 2024

    2024

  24. [30]

    Deep unsupervised learning using nonequi- librium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequi- librium thermodynamics,” in International conference on machine learning. PMLR, 2015, pp. 2256–2265

    2015

  25. [31]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems, vol. 33, pp. 6840–6851, 2020

    2020

  26. [32]

    Denoising diffu- sion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffu- sion implicit models,” arXiv preprint arXiv:2010.02502 , 2020

    2010 arXiv

  27. [33]

    Ilvr: Conditioning method for denoising diffusion probabilis- tic models,

    J. Choi, S. Kim, Y . Jeong, Y . Gwon, and S. Yoon, “Ilvr: Conditioning method for denoising diffusion probabilis- tic models,” arXiv preprint arXiv:2108.02938 , 2021

    2021 arXiv

  28. [34]

    An overview of image generation of industrial surface defects,

    X. Zhong, J. Zhu, W. Liu, C. Hu, Y . Deng, and Z. Wu, “An overview of image generation of industrial surface defects,” Sensors, vol. 23, no. 19, p. 8160, 2023

    2023

  29. [36]

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

    R. Gal, Y . Alaluf, Y . Atzmon, O. Patashnik, A. H. Bermano, G. Chechik, and D. Cohen-Or, “An im- age is worth one word: Personalizing text-to-image generation using textual inversion,” arXiv preprint arXiv:2208.01618, 2022

    2022 arXiv

  30. [37]

    Dreambooth: Fine tuning text-to- image diffusion models for subject-driven generation,

    N. Ruiz, Y . Li, V . Jampani, Y . Pritch, M. Rubinstein, and K. Aberman, “Dreambooth: Fine tuning text-to- image diffusion models for subject-driven generation,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , 2023, pp. 22 500–22 510

    2023

  31. [38]

    Lora: Low-rank adaptation of large language models

    E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, W. Chen et al., “Lora: Low-rank adaptation of large language models.” ICLR, vol. 1, no. 2, p. 3, 2022

    2022

  32. [39]

    Instantid: Zero-shot identity-preserving generation in seconds,

    Q. Wang, X. Bai, H. Wang, Z. Qin, A. Chen, H. Li, X. Tang, and Y . Hu, “Instantid: Zero-shot identity-preserving generation in seconds,”arXiv preprint arXiv:2401.07519, 2024

    2024 arXiv

  33. [40]

    Advancing pose-guided image synthesis with progressive conditional diffusion models,

    F. Shen, H. Ye, J. Zhang, C. Wang, X. Han, and W. Yang, “Advancing pose-guided image synthesis with progressive conditional diffusion models,” arXiv preprint arXiv:2310.06313, 2023

    2023 arXiv

  34. [41]

    Boosting consistency in story visualization with rich-contextual conditional diffusion models,

    F. Shen, H. Ye, S. Liu, J. Zhang, C. Wang, X. Han, and Y . Wei, “Boosting consistency in story visualization with rich-contextual conditional diffusion models,” in Proceedings of the AAAI Conference on Artificial Intel- ligence, vol. 39, no. 7, 2025, pp. 6785–6794

    2025

  35. [42]

    Auto-encoding varia- tional bayes,

    D. P. Kingma, M. Welling et al., “Auto-encoding varia- tional bayes,” 2013

    2013

  36. [43]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al., “Learning transferable visual models from natural language supervision,” in International conference on machine learning. PmLR, 2021, pp. 8748–8763

    2021

  37. [45]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,” Advances in neural information processing systems , vol. 30, 2017

    2017

  38. [46]

    A note on the inception score,

    S. Barratt and R. Sharma, “A note on the inception score,” arXiv preprint arXiv:1801.01973 , 2018

    2018 arXiv

  39. [47]

    Dcgan: Deep convolutional gan with attention module for remote view classification,

    A. Patil et al. , “Dcgan: Deep convolutional gan with attention module for remote view classification,” in 2021 International Conference on Forensics, Analytics, Big Data, Security (FABS), vol. 1. IEEE, 2021, pp. 1–10

    2021

  40. [48]

    Alias-free generative adversar- ial networks,

    T. Karras, M. Aittala, S. Laine, E. H ¨ark¨onen, J. Hellsten, J. Lehtinen, and T. Aila, “Alias-free generative adversar- ial networks,” in Proc. NeurIPS, 2021

    2021

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.