Full search page →

IKC: iterative kernel correction (CVPR 2019)

Optional · paper guide 2 of 11 for DLCV 09A.1 · Blind and Real-World Image Super-Resolution. Citation numbers refer to the reference list at the bottom of this page.

Navigate: this page and this track: Navigate: this page and this track

On this page

Supplementary notes · 3

  1. The high-order degradation pipeline, step by step
  2. Sinc filters and ringing
  3. Why the kernel is hard to identify

Paper guides · 11

  1. SRMD: one network for many known degradations (CVPR 2018)
  2. IKC: iterative kernel correction (CVPR 2019)
  3. KernelGAN: kernel estimation with an internal GAN (NeurIPS 2019)
  4. DAN: unfolding the alternating optimization (NeurIPS 2020)
  5. DASR: contrastive degradation representations (CVPR 2021)
  6. BSRGAN: a practical degradation model (ICCV 2021)
  7. Real-ESRGAN: high-order degradation with pure synthetic data (ICCVW 2021)
  8. StableSR: a time-aware encoder for a frozen diffusion prior (IJCV 2024)
  9. DiffBIR: remove the degradation, then regenerate (ECCV 2024)
  10. SeeSR: degradation-aware semantic prompts (CVPR 2024)
  11. SUPIR: scaling up the prior, the data and the text (CVPR 2024)

Deep Learning for Computer Vision

  1. 02 · Deep Image Classification
View the whole course tree →

Challenge

Non-blind networks such as SRMD work well when given the right kernel, but a predicted kernel is never exactly right, and small errors are magnified into visible artifacts. Gu et al. observe that these artifacts are regular: if the kernel used for SR is sharper than the true one, the result is over-smoothed; if it is smoother, the result shows ringing [8]. Can a network learn to read this signature and correct the kernel?

Method

The predictor initializes a kernel code, while the corrector uses the current super-resolved image to estimate an update; correction is iterative rather than a single oracle kernel input.
Original paper figure 5 — The predictor initializes a kernel code, while the corrector uses the current super-resolved image to estimate an update; correction is iterative rather than a single oracle kernel input.

Figure source: [8], Published paper PDF, PDF page 5. Original PDF.

Problem setting. ILR=(k⊗IHR)↓s+nI^{LR} = (k \otimes I^{HR})\downarrow_s + n with isotropic Gaussian kernels and bicubic downsampling; a noisy version uses Gaussian noise. Kernels are 21×2121 \times 21 and, as in SRMD, are reduced by a PCA matrix M∈Rb×l2M \in \mathbb R^{b \times l^2} to a code h=Mk∈Rbh = M k \in \mathbb R^b, where l=21l = 21 is the kernel size and bb the code length.

Three networks.

  1. SR model FF (SFTMD). An SRResNet-style network [70] with spatial feature transform (SFT) layers [32] after every convolution in the residual blocks and after the global residual connection. The kernel code is stretched into maps HH of size b×H×Wb \times H \times W; for feature maps FF of size Cf×H×WC_f \times H \times W, a small CNN takes the concatenation of FF and HH and outputs a scale γ\gamma and a shift β\beta, and the layer outputs SFT(F,H)=γ⊙F+β\mathrm{SFT}(F, H) = \gamma \odot F + \beta, with ⊙\odot the elementwise product. Unlike SRMD’s input concatenation, the kernel influences every layer, and the image features are modulated rather than mixed with constant maps.
  2. Predictor PP. Four convolution layers with leaky ReLU and global average pooling give an initial code h0=P(ILR)h_0 = P(I^{LR}), trained with θP=arg⁡min⁡∥k−P(ILR;θP)∥22\theta_P = \arg\min \| k - P(I^{LR}; \theta_P) \|_2^2 (in code space).
  3. Corrector CC. It takes the current SR image ISRI^{SR} and the current code hh. Five convolution layers extract features of ISRI^{SR} (where blur or ringing is visible); two fully connected layers process hh, which is then stretched into maps; the two are concatenated and three 1×11 \times 1 convolutions plus global average pooling output a correction Δh\Delta h. It is trained so that h+C(ISR,h)h + C(I^{SR}, h) approaches the true code.

Algorithm. h0=P(ILR)h_0 = P(I^{LR}), I0SR=F(ILR,h0)I_0^{SR} = F(I^{LR}, h_0); then for i=1,…,ti = 1, \dots, t: Δhi=C(Ii−1SR,hi−1)\Delta h_i = C(I_{i-1}^{SR}, h_{i-1}), hi=hi−1+Δhih_i = h_{i-1} + \Delta h_i, IiSR=F(ILR,hi)I_i^{SR} = F(I^{LR}, h_i). Small corrections repeated many times avoid the over- or under-correction of a single step.

Training. SFTMD is pretrained with MSE on synthetic pairs (kernel widths in [0.2,2][0.2, 2], [0.2,3][0.2, 3], [0.2,4][0.2, 4] for ×2,3,4\times 2, 3, 4; 3,450 DIV2K and Flickr2K images [47], [71]); it is then frozen, and PP and CC are trained alternately with t=7t = 7 iterations, Adam [64] and learning rate 10−410^{-4}.

Gaussian8 benchmark. To compare blind methods on known kernels, the paper defines eight isotropic kernels per scale with widths in [0.80,1.60][0.80, 1.60], [1.35,2.40][1.35, 2.40] and [1.80,3.20][1.80, 3.20] for ×2\times 2, ×3\times 3, ×4\times 4, applied before bicubic downsampling of Set5, Set14, BSD100, Urban100 and Manga109 [72].

Results

On Gaussian8 at ×4\times 4, IKC reaches 31.52 dB on Set5, 28.26 on Set14, 27.29 on BSD100, 25.33 on Urban100 and 29.90 on Manga109, against 26.45, 24.78, 24.97, 22.11 and 23.53 dB for ZSSR and 24.57, 22.79, 23.29, 20.35 and 21.50 dB for bicubic interpolation. Running the SR model with only the predictor’s initial kernel (“P + SFTMD”) gives 29.29 dB on Set5 at ×4\times 4, so correction adds 2.23 dB there. At ×2\times 2 the Set5 result is 36.62 dB. Deblurring the LR image with a blind deblurring method before a bicubic-trained SR network performs far worse (18.10 dB on Set5 at ×4\times 4), and deblurring after SR is better but still below IKC (28.69 dB). With a direct downsampler that IKC was not trained on, it still leads on BSD100 (27.06 dB at kernel width 2.0 in the paper’s Table 3). On real images (a historic photo, “Chip”) IKC gives artifact-free results without manual kernel search.

Ablations

  • SFT versus concatenation (non-blind, ground-truth kernels). On Set5 ×4\times 4 with width 1.3, SFTMD gets 32.39 dB, SRMDNF 32.00, the same SRResNet backbone with SFT layers replaced by direct concatenation 31.69, and the backbone with the kernel concatenated only at the first layer 29.11 dB. In this architecture and experiment, repeated affine modulation preserves the conditioning better than these concatenation variants.
  • Number of iterations. On the intermediate results of one image, PSNR rises from 23.60 dB (initial prediction) to 25.65, 27.75, 27.94 and 27.97 dB after one to four corrections; a curve on Set14 shows PSNR and SSIM increasing and then converging. One correction step is not enough.
  • PCA codes versus kernel widths. Learning the PCA code instead of the Gaussian width generalizes better to the unseen direct downsampler: 27.06 versus 26.85 dB on BSD100 at width 2.0, and 26.98 versus 26.75 dB at width 3.0.

Conclusion

Kernel mismatch leaves regular artifacts, and a network can use them to correct the kernel iteratively. Together with SFT-based conditioning this made blind SR with Gaussian kernels several dB better than earlier blind methods.

Problems

  • Isotropic Gaussians only. The authors note that real kernels include slight motion blur and leave anisotropic kernels to future work.
  • Three separately trained networks and seven SR passes make IKC slow (about 3.93 s per image in DAN’s measurements) and the predictor, corrector and SR model are never optimized jointly. DAN unfolds the alternation into one end-to-end network.
  • The SR model is trained with true kernels but used with estimated ones; DAN trains the restorer on estimates instead.
  • No JPEG or processing: like SRMD, IKC stays inside the classical model.

References

  1. N. Efrat, D. Glasner, A. Apartsin, B. Nadler and A. Levin, “Accurate Blur Models vs. Image Priors in Single Image Super-Resolution,” in Proc. IEEE ICCV, 2013, pp. 2832–2839. doi · CVF
  2. K. Zhang, W. Zuo and L. Zhang, “Learning a Single Convolutional Super-Resolution Network for Multiple Degradations,” CVPR, 2018. paper
  3. A. Liu, Y. Liu, J. Gu, Y. Qiao, C. Dong, “Blind Image Super-Resolution: A Survey and Beyond,” IEEE TPAMI (early access), 2022. source
  4. Honggang Chen, Xiaohai He, Linbo Qing, Yuanyuan Wu, Chao Ren, Ray E. Sheriff, Ce Zhu, “Real-world single image super-resolution: A brief review,” Information Fusion, 2022. source
  5. B. B. Moser, A. S. Shanbhag, F. Raue, S. Frolov, S. Palacio and A. Dengel, “Diffusion Models, Image Super-Resolution, and Everything: A Survey,” IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 7, pp. 11793–11813, 2025. doi · arXiv
  6. Y. Xu, S. R. Tseng, Y. Tseng, H. Kuo, Y. Tsai, “Unified Dynamic Convolutional Network for Super-Resolution with Variational Degradations,” CVPR, 2020. source
  7. K. Zhang, L. Van Gool, R. Timofte, “Deep Unfolding Network for Image Super-Resolution,” CVPR, 2020. source
  8. J. Gu, H. Lu, W. Zuo and C. Dong, “Blind Super-Resolution With Iterative Kernel Correction,” CVPR, 2019. paper
  9. Z. Luo, Y. Huang, S. Li, L. Wang and T. Tan, “Unfolding the Alternating Optimization for Blind Super Resolution,” NeurIPS, 2020. paper
  10. Z. Luo, H. Huang, L. Yu, Y. Li, H. Fan and S. Liu, “Deep Constrained Least Squares for Blind Image Super-Resolution,” CVPR, 2022. paper
  11. T. Michaeli, M. Irani, “Nonparametric Blind Super-resolution,” Proceedings of the IEEE International Conference on Computer Vision, 2013. source
  12. S. Bell-Kligler, A. Shocher and M. Irani, “Blind Super-Resolution Kernel Estimation using an Internal-GAN,” NeurIPS, 2019. paper
  13. A. Shocher, N. Cohen and M. Irani, “‘Zero-Shot’ Super-Resolution Using Deep Internal Learning,” in Proc. IEEE/CVF CVPR, 2018, pp. 3118–3126. doi · arXiv
  14. L. Wang, Y. Wang, X. Dong, Q. Xu, J. Yang, W. An and Y. Guo, “Unsupervised Degradation Representation Learning for Blind Super-Resolution,” CVPR, 2021. paper
  15. B. Xia, Y. Zhang, Y. Wang, Y. Tian, W. Yang, R. Timofte and L. Van Gool, “Knowledge Distillation Based Degradation Estimation for Blind Super-Resolution,” ICLR, 2023. paper
  16. K. Zhang, J. Liang, L. Van Gool, R. Timofte, “Designing a Practical Degradation Model for Deep Blind Image Super-Resolution,” Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021. source
  17. X. Wang, L. Xie, C. Dong, Y. Shan, “Real-ESRGAN: Training Real-World Blind Super-Resolution With Pure Synthetic Data,” Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021. source
  18. J. Liang, H. Zeng, L. Zhang, “Efficient and Degradation-Adaptive Network for Real-World Image Super-Resolution,” ECCV, 2022. source
  19. J. Cai, H. Zeng, H. Yong, Z. Cao, L. Zhang, “Toward Real-World Single Image Super-Resolution: A New Benchmark and a New Model,” Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019. source
  20. P. Wei, Z. Xie, H. Lu, Z. Zhan, Q. Ye, W. Zuo, L. Lin, “Component Divide-and-Conquer for Real-World Image Super-Resolution,” ECCV, 2020. source
  21. C. Chen, Z. Xiong, X. Tian, Z. Zha, F. Wu, “Camera Lens Super-Resolution,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019. source
  22. Y. Yuan, S. Liu, J. Zhang, Y. Zhang, C. Dong, L. Lin, “Unsupervised Image Super-Resolution using Cycle-in-Cycle Generative Adversarial Networks,” CVPR Workshops, 2018. source
  23. A. Bulat, J. Yang, G. Tzimiropoulos, “To learn image super-resolution, use a GAN to learn how to do image degradation first,” ECCV, 2018. source
  24. M. Fritsche, S. Gu, R. Timofte, “Frequency Separation for Real-World Super-Resolution,” ICCV Workshops, 2019. source
  25. X. Ji, Y. Cao, Y. Tai, C. Wang, J. Li, F. Huang, “Real-World Super-Resolution via Kernel Estimation and Noise Injection,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020. source
  26. J. Wang, Z. Yue, S. Zhou, K. C. K. Chan, C. C. Loy, “Exploiting Diffusion Prior for Real-World Image Super-Resolution,” IJCV, 2024. source
  27. X. Lin, J. He, Z. Chen, et al., “DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior,” ECCV, 2024. source
  28. T. Yang, R. Wu, P. Ren, X. Xie, L. Zhang, “Pixel-Aware Stable Diffusion for Realistic Image Super-resolution and Personalized Stylization,” ECCV, 2024. source
  29. R. Wu, T. Yang, L. Sun, Z. Zhang, S. Li, L. Zhang, “SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024. source · supplementary experiments
  30. F. Yu, J. Gu, Z. Li, et al., “Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic Image Restoration In the Wild,” CVPR, 2024. source
  31. Z. Yue, J. Wang, C. C. Loy, “ResShift: Efficient Diffusion Model for Image Super-resolution by Residual Shifting,” NeurIPS, 2023. source
  32. X. Wang, K. Yu, C. Dong, C. C. Loy, “Recovering Realistic Texture in Image Super-Resolution by Deep Spatial Feature Transform,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2018. source
  33. K. He, H. Fan, Y. Wu, S. Xie, R. Girshick, “Momentum Contrast for Unsupervised Visual Representation Learning,” CVPR, 2020. source
  34. X. Wang, K. Yu, S. Wu, J. Gu, Y. Liu, C. Dong, Y. Qiao and C. C. Loy, “ESRGAN: Enhanced Super-Resolution Generative Adversarial Networks,” in Proc. ECCV Workshops, 2018 (Lecture Notes in Computer Science, 2019), pp. 63–79. doi · arXiv
  35. O. Ronneberger, P. Fischer and T. Brox, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” in Proc. MICCAI, 2015, pp. 234–241. doi · arXiv
  36. T. Miyato, T. Kataoka, M. Koyama, Y. Yoshida, “Spectral Normalization for Generative Adversarial Networks,” ICLR, 2018. source
  37. J. Liang, H. Zeng, L. Zhang, “Details or Artifacts: A Locally Discriminative Learning Approach to Realistic Image Super-Resolution,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022. source
  38. A. Ignatov, N. Kobyshev, R. Timofte, K. Vanhoey, L. Van Gool, “DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks,” ICCV, 2017. source
  39. A. Lugmayr, M. Danelljan, R. Timofte, “Unsupervised Learning for Real-World Super-Resolution,” arXiv preprint, 2019. source
  40. A. Lugmayr, M. Danelljan, R. Timofte, et al., “NTIRE 2020 Challenge on Real-World Image Super-Resolution: Methods and Results,” CVPR Workshops, 2020. source
  41. R. Rombach, A. Blattmann, D. Lorenz, P. Esser, B. Ommer, “High-Resolution Image Synthesis with Latent Diffusion Models,” CVPR, 2022. source
  42. L. Zhang, A. Rao, M. Agrawala, “Adding Conditional Control to Text-to-Image Diffusion Models,” ICCV, 2023. source
  43. Y. Ai, X. Zhou, H. Huang, X. Han, Z. Chen, Q. You, H. Yang, “DreamClear: High-Capacity Real-World Image Restoration with Privacy-Safe Dataset Curation,” NeurIPS, 2024. source
  44. Z. Duan, J. Zhang, X. Jin, et al., “DiT4SR: Taming Diffusion Transformer for Real-World Image Super-Resolution,” ICCV, 2025. source
  45. R. Wu, L. Sun, Z. Ma, L. Zhang, “One-Step Effective Diffusion Network for Real-World Image Super-Resolution,” NeurIPS, 2024. source
  46. Y. Wang, W. Yang, X. Chen, et al., “SinSR: Diffusion-Based Image Super-Resolution in a Single Step,” CVPR, 2024. source
  47. E. Agustsson and R. Timofte, “NTIRE 2017 Challenge on Single Image Super-Resolution: Dataset and Study,” in Proc. IEEE CVPR Workshops, 2017, pp. 1122–1131. doi · CVF
  48. Z. Wang, A. C. Bovik, H. R. Sheikh and E. P. Simoncelli, “Image Quality Assessment: From Error Visibility to Structural Similarity,” IEEE Transactions on Image Processing, vol. 13, no. 4, pp. 600–612, 2004. doi
  49. R. Zhang, P. Isola, A. A. Efros, E. Shechtman and O. Wang, “The Unreasonable Effectiveness of Deep Features as a Perceptual Metric,” in Proc. IEEE/CVF CVPR, 2018, pp. 586–595. doi · arXiv
  50. K. Ding, K. Ma, S. Wang, E. P. Simoncelli, “Image Quality Assessment: Unifying Structure and Texture Similarity,” IEEE TPAMI, 2022. source
  51. A. Mittal, R. Soundararajan and A. C. Bovik, “Making a “Completely Blind” Image Quality Analyzer,” IEEE Signal Processing Letters, vol. 20, no. 3, pp. 209–212, 2013. doi
  52. J. Ke, Q. Wang, Y. Wang, P. Milanfar, F. Yang, “MUSIQ: Multi-scale Image Quality Transformer,” ICCV, 2021. source
  53. S. Yang, T. Wu, S. Shi, S. Lao, Y. Gong, M. Cao, J. Wang, Y. Yang, “MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment,” CVPR Workshops, 2022. source
  54. J. Wang, K. C. K. Chan, C. C. Loy, “Exploring CLIP for Assessing the Look and Feel of Images,” AAAI, 2023. source
  55. M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, S. Hochreiter, “GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium,” NeurIPS, 2017. source
  56. R. Timofte, S. Gu, J. Wu, L. Van Gool, “NTIRE 2018 Challenge on Single Image Super-Resolution: Methods and Results,” Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2018. source
  57. J. Cai, S. Gu, R. Timofte, L. Zhang, “NTIRE 2019 Challenge on Real Image Super-Resolution: Methods and Results,” Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2019. source
  58. A. Lugmayr, M. Danelljan, R. Timofte, et al., “AIM 2019 Challenge on Real-World Image Super-Resolution: Methods and Results,” ICCV Workshops, 2019. source
  59. P. Wei, H. Lu, R. Timofte, et al., “AIM 2020 Challenge on Real Image Super-Resolution: Methods and Results,” ECCV Workshops, 2020. source
  60. Z. Chen, K. Liu, J. Gong, J. Wang, L. Sun, Z. Wu, R. Timofte, Y. Zhang et al., “NTIRE 2025 Challenge on Image Super-Resolution (×4): Methods and Results,” in Proc. IEEE/CVF CVPR Workshops, 2025, pp. 1525–1535. doi · arXiv
  61. Le Zhang, Ao Li, Qibin Hou, Ce Zhu, Yonina C. Eldar, “Deep-Learning-Empowered Super Resolution: A Comprehensive Survey and Future Prospects,” Proceedings of the IEEE, 2025. source
  62. W. Ren, R. Goyal, Z. Hu, T. T. Aumentado-Armstrong, I. Mohomed and A. Levinshtein, “Hallucination Score: Towards Mitigating Hallucinations in Generative Image Super-Resolution,” arXiv:2507.14367, 2025. arXiv
  63. Real-ESRGAN contributors, “Real-ESRGAN model and ×4-plus training configuration,” official source, accessed 2026-11-02. model · configuration
  64. D. P. Kingma, J. Ba, “Adam: A Method for Stochastic Optimization,” ICLR, 2015. source
  65. M. Bevilacqua, A. Roumy, C. Guillemot and M.-L. Alberi Morel, “Low-Complexity Single-Image Super-Resolution based on Nonnegative Neighbor Embedding,” in Proc. BMVC, 2012, pp. 135.1–135.10. doi
  66. R. Zeyde, M. Elad and M. Protter, “On Single Image Scale-Up Using Sparse-Representations,” in Curves and Surfaces, Lecture Notes in Computer Science, Springer, 2012, pp. 711–730. doi
  67. D. Martin, C. Fowlkes, D. Tal and J. Malik, “A Database of Human Segmented Natural Images and Its Application to Evaluating Segmentation Algorithms and Measuring Ecological Statistics,” in Proc. IEEE ICCV, 2001, vol. 2, pp. 416–423. doi
  68. J.-B. Huang, A. Singh and N. Ahuja, “Single Image Super-Resolution from Transformed Self-Exemplars,” in Proc. IEEE CVPR, 2015, pp. 5197–5206. doi · CVF
  69. J. Kim, J. K. Lee and K. M. Lee, “Accurate Image Super-Resolution Using Very Deep Convolutional Networks,” in Proc. IEEE CVPR, 2016, pp. 1646–1654. doi · arXiv
  70. C. Ledig, L. Theis, F. Huszár, J. Caballero, A. Cunningham, A. Acosta, A. Aitken, A. Tejani, J. Totz, Z. Wang and W. Shi, “Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network,” in Proc. IEEE CVPR, 2017, pp. 105–114. doi · arXiv
  71. R. Timofte, E. Agustsson, L. Van Gool, M.-H. Yang, L. Zhang et al., “NTIRE 2017 Challenge on Single Image Super-Resolution: Methods and Results,” in Proc. IEEE CVPR Workshops, 2017, pp. 1110–1121. doi · CVF
  72. Y. Matsui, K. Ito, Y. Aramaki, A. Fujimoto, T. Ogawa, T. Yamasaki and K. Aizawa, “Sketch-Based Manga Retrieval Using Manga109 Dataset,” Multimedia Tools and Applications, vol. 76, no. 20, pp. 21811–21838, 2017. doi
  73. KernelGAN contributors, “Adversarial losses and kernel constraints,” official source, accessed 2026-11-02. loss definitions · alternating updates
  74. X. Mao, Q. Li, H. Xie, R. Y. K. Lau, Z. Wang, S. P. Smolley, “Least Squares Generative Adversarial Networks,” ICCV, 2017. source
  75. DASR contributors, “MoCo logits and contrastive cross-entropy implementation,” official DASR source, commit 51524af8021c94a502ee002e135867a1b75dc7d0, accessed 2026-11-02. encoder · loss
  76. Y. Zhang, K. Li, K. Li, L. Wang, B. Zhong and Y. Fu, “Image Super-Resolution Using Very Deep Residual Channel Attention Networks,” in Proc. ECCV, 2018, pp. 294–310. doi · arXiv
  77. K. Simonyan and A. Zisserman, “Very Deep Convolutional Networks for Large-Scale Image Recognition,” in Proc. ICLR, 2015. arXiv
  78. Y. Zhang, X. Huang, J. Ma, et al., “Recognize Anything: A Strong Image Tagging Model,” CVPR Workshops, 2024. source
  79. E. J. Hu, Y. Shen, P. Wallis, Z. Allen-Zhu, Y. Li, S. Wang, L. Wang, W. Chen, “LoRA: Low-Rank Adaptation of Large Language Models,” ICLR, 2022. source
  80. D. Podell, Z. English, K. Lacey, A. Blattmann, T. Dockhorn, J. Müller, J. Penna, R. Rombach, “SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis,” ICLR, 2024. source
  81. H. Liu, C. Li, Q. Wu, Y. J. Lee, “Visual Instruction Tuning,” NeurIPS, 2023. source
  82. T. Karras, M. Aittala, T. Aila, S. Laine, “Elucidating the Design Space of Diffusion-Based Generative Models,” NeurIPS, 2022. source