請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81346完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 吳家麟(Ja-Ling Wu) | |
| dc.contributor.author | Chien-Hung Lin | en |
| dc.contributor.author | 林建宏 | zh_TW |
| dc.date.accessioned | 2022-11-24T03:44:34Z | - |
| dc.date.available | 2021-08-06 | |
| dc.date.available | 2022-11-24T03:44:34Z | - |
| dc.date.copyright | 2021-08-06 | |
| dc.date.issued | 2021 | |
| dc.date.submitted | 2021-07-19 | |
| dc.identifier.citation | Rameen, A., Yipeng, Q., Peter, W. Image2stylegan: How to embed images into the stylegan latent space? ICCV 2019 Jiapeng Zhu, Yujun Shen, Deli Zhao, Bolei Zhou. In-Domain GAN Inversion for Real Image Editing. ECCV 2020 Jinjin Gu, Yujun Shen, and Bolei Zhou. Image processing using multi-code gan prior. CVPR, 2020. Y. Shen, J. Gu, X. Tang, and B. Zhou. Interpreting the latent space of gans for semantic face editing. CVPR 2020 E. Hark onen, A. Hertzmann, J. Lehtinen, and S. Paris. Ganspace: Discovering interpretable gan controls. NeurIPS 2020 Yujun Shen, Bolei Zhou. Closed-Form Factorization of Latent Semantics in GANs. CVPR 2021 Y.-L. Pan, J.-C. Chen, and J.-L. Wu. A Multi-factor Combinations Enhanced Reversible Privacy Protection System for Facial Images. In IEEE International Conference on Multimedia and Expo (ICME), IEEE, 2021. 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. Proc. NIPS, pages 6626–6637, 2017. Wang, Z. and Sheikh, H.R. Image Quality Assessment: From Error Visibility to Structural Similarity. 2004 IEEE Transactions on Image Processing, 13, No. 4. Richard Zhang, Phillip Isola, Alexei A. Efros, Eli Shechtman, Oliver Wang. The Unreasonable Effectiveness of Deep Features as a Perceptual Metric. CVPR, 2018. Ziwei Liu, Ping Luo, Xiaogang Wang, Xiaoou Tang. Large-scale CelebFaces Attributes (CelebA) Dataset. Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, Yoshua Bengio. Generative Adversarial Networks. NIPS, 2014. Jun-Yan Zhu, Taesung Park, Phillip Isola, Alexei A. Efros. Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks. ICCV 2017 Tero Karras, Samuli Laine, Timo Aila. A Style-Based Generator Architecture for Generative Adversarial Networks. CVPR 2019 Taesung Park, Ming-Yu Liu, Ting-Chun Wang, Jun-Yan Zhu. Semantic Image Synthesis with Spatially-Adaptive Normalization. CVPR 2019 oral paper. Sudipto Mukherjee, Himanshu Asnani, Eugene Lin, Sreeram Kannan. ClusterGAN : Latent Space Clustering in Generative Adversarial Networks. CoRR, abs/1809.03627, 2018. Xun Huang, Serge Belongie. Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization. ICCV 2017 | |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/81346 | - |
| dc.description.abstract | "現今機器學習在圖像生成的技術已越來越成熟，其中又以 styleGAN、BigGAN 等對抗式生成網路的結果更為令人驚豔。但不幸地，這些模型的架構讓我們很難對輸出圖像進行調控，造成徒有生成結果好，可再細部操作地方卻很少的窘進。 因此就有人嘗試在隱藏空間 (Latent Space) 中對隱藏碼 (latent Code) 進行編輯，以達到可在不改變原模型的架構及已耗時耗能學好的參數為前提下，僅透過將新的隱藏碼嵌入原模型中，就做到對輸出圖像進行調控的效果。但這些方法都存有不同的限制：例如，不適用於隱藏空間較大的狀況或會產生不同特徵間相互糾纏的問題。 因此在本論文中，我們提出兩種方法來解決上述所提及的問題：一種是透過將原隱藏空間壓縮來讓受限於隱藏空間大小的分析方法能重新起用;另一種是透過額外訓練一個較簡單之模型來針對不同的隱藏碼產生出最適合調控效果的新隱藏碼。這兩種方法相比於之前的方法能適用於更多種類的模型,且仍然能做到圖片控制的效果。" | zh_TW |
| dc.description.provenance | Made available in DSpace on 2022-11-24T03:44:34Z (GMT). No. of bitstreams: 1 U0001-1607202115185100.pdf: 6055082 bytes, checksum: c387edf1a6e9dff25b94a467278c8761 (MD5) Previous issue date: 2021 | en |
| dc.description.tableofcontents | 口試委員會審定書 i 誌謝 iii 中文摘要 iv ABSTRACT v List of Figures viii List of Tables x Chapter 1. Introduction 1 Chapter 2. Related Works 3 2.1 Generative Adversarial Network 3 2.2 Latent Space Research 5 2.2.1 Image Inversion to Latent Space 5 2.2.2 Latent Code Manipulation 8 Chapter 3. Proposed Method for Facial Image Manipulation 10 3.1 Preliminary 10 3.2 Supervised Method 10 3.3 Semi-Supervised Method 15 Chapter 4. Experiments 21 4.1 Dataset Metrics 21 4.2 Auto Encoder – Train by own self 22 4.2.1 Supervised Method 22 4.2.2 Semi-Supervised Method 26 4.2.3 Quantitative Comparison 29 4.3 MfM - Privacy Protection System for Facial Image 33 4.3.1 Supervised Method 33 4.3.2 Semi-Supervised Method 36 4.3.3 Quantitative Comparison 39 4.4 Comparison 42 Chapter 5. Discussion 43 Chapter 6. Conclusions Future Works 44 References 45 Appendix 47 A. Multiple attribute control 47 B. Network Details 51 | |
| dc.language.iso | en | |
| dc.subject | 對抗式生成網路 | zh_TW |
| dc.subject | 隱藏空間 | zh_TW |
| dc.subject | 圖片控制 | zh_TW |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | Generative Adversarial Network | en |
| dc.subject | Deep Learning | en |
| dc.subject | Image Manipulation | en |
| dc.subject | Latent Space | en |
| dc.title | 基於屬性潛在空間上的面部圖像處理 | zh_TW |
| dc.title | Attribute-Based Facial Image Manipulation on Latent Space | en |
| dc.date.schoolyear | 109-2 | |
| dc.description.degree | 碩士 | |
| dc.contributor.oralexamcommittee | 周承復(Hsin-Tsai Liu),張嘉淵(Chih-Yang Tseng),陳俊良,陳駿丞 | |
| dc.subject.keyword | 隱藏空間,圖片控制,深度學習,對抗式生成網路, | zh_TW |
| dc.subject.keyword | Latent Space,Image Manipulation,Deep Learning,Generative Adversarial Network, | en |
| dc.relation.page | 51 | |
| dc.identifier.doi | 10.6342/NTU202101513 | |
| dc.rights.note | 同意授權(限校園內公開) | |
| dc.date.accepted | 2021-07-19 | |
| dc.contributor.author-college | 電機資訊學院 | zh_TW |
| dc.contributor.author-dept | 資訊工程學研究所 | zh_TW |
| 顯示於系所單位： | 資訊工程學系 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| U0001-1607202115185100.pdf 授權僅限NTU校內IP使用（校園外請利用VPN校外連線服務） | 5.91 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。