PAST SIFATLI YUZ TASVIRLARINI GFPGAN VA CODEFORMER ALGORITMLARI YORDAMIDA YAXSHILASH: QIYOSIY TAHLIL
Keywords:
yuz tasviri, tasvirni tiklash, tasvir sifatini yaxshilash, GFPGAN, CodeFormer, SSIM, LPIPS.Abstract
Ushbu ishda GFPGAN v1.4 va CodeFormer algoritmlari past sifatli yuz tasvirini yaxshilash vazifasida amaliy sinab ko‘rildi. Bir xil past sifatli tasvir har ikki algoritm yordamida alohida qayta ishlandi. Olingan natijalar asl sifatli tasvir bilan vizual tarzda, shuningdek SSIM va LPIPS ko‘rsatkichlari asosida solishtirildi. Tajribadan asosiy maqsad ushbu ikki algoritm yuz tasvirini yaxshilashda amalda qanday natija berishini ko‘rib chiqish va ular orasidagi farqlarni tahlil qilishdan iborat. Natijalarda ikkala algoritm ham yuzning ayrim lokal detallarini aniqroq ko‘rsatdi. LPIPS qiymati GFPGAN va CodeFormer natijalarida boshlang‘ich past sifatli tasvirga nisbatan kamaydi, SSIM qiymatida esa sezilarli o‘sish kuzatilmadi. CodeFormer ushbu tajribada biroz pastroq LPIPS qiymatini ko‘rsatdi, ammo ikki algoritm orasidagi farq juda kichik bo‘ldi. Natijalar yuz tasvirini tiklashni faqat vizual ko‘rinish bilan emas, strukturaviy va perceptual ko‘rsatkichlar bilan birga baholash foydali ekanini ko‘rsatdi.
References
Wang X., Li Y., Zhang H., Shan Y. Towards Real-World Blind Face Restoration with Generative Facial Prior. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 9168–9178.
Zhou S., Chan K. C. K., Li C., Loy C. C. Towards Robust Blind Face Restoration with Codebook Lookup Transformer. Advances in Neural Information Processing Systems (NeurIPS), Vol. 35, 2022, pp. 30599–30611.
Wang Z., Bovik A. C., Sheikh H. R., Simoncelli E. P. Image Quality Assessment: From Error Visibility to Structural Similarity. IEEE Transactions on Image Processing, Vol. 13, No. 4, 2004, pp. 600–612.
Zhang R., Isola P., Efros A. A., Shechtman E., Wang O. The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 586–595.
Li X., Chen C., Zhou S., Lin X., Zuo W., Zhang L. Blind Face Restoration via Deep Multi-Scale Component Dictionaries. Proceedings of the European Conference on Computer Vision (ECCV), 2020. DOI: 10.1007/978-3-030-58545-7_23.
Yang L., Liu C., Wang P., Wang S., Ren P., Ma S., Gao W. HiFaceGAN: Face Renovation via Collaborative Suppression and Replenishment. Proceedings of the 28th ACM International Conference on Multimedia (ACM MM), 2020, pp. 1551–1560.
Yang T., Ren P., Xie X., Zhang L. GAN Prior Embedded Network for Blind Face Restoration in the Wild. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 672–681.
Wang Z., Zhang J., Chen R., Wang W., Luo P. RestoreFormer: High-Quality Blind Face Restoration from Undegraded Key-Value Pairs. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 17512–17521.
Gu Y., Wang X., Xie L., Dong C., Li G., Shan Y., Cheng M.-M. VQFR: Blind Face Restoration with Vector-Quantized Dictionary and Parallel Decoder. Proceedings of the European Conference on Computer Vision (ECCV), 2022, pp. 126–143.
Menon S., Damian A., Hu S., Ravi N., Rudin C. PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020, pp. 2437–2445.
Downloads
Published
How to Cite
Issue
Section
License

This work is licensed under a Creative Commons Attribution 4.0 International License.
Articles published in IQRO — Scientific Methodological Journal are licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0).
Authors retain copyright and grant the journal the right of first publication.
Users may read, download, copy, distribute, print, search, or link to the full texts of articles, provided that the original authors and the journal are properly cited.
Published materials may be used, distributed, and reproduced in any medium, provided that the original work is properly cited.



Uzbekistan
Türkiye
Kazakhstan
Kyrgyzstan
Turkmenistan
Azerbaijan
India
Republic of Korea
Japan
United States of America
Kosovo