0% Complete
صفحه اصلی
/
سی و دومین کنفرانس بین المللی مهندسی برق
Enhancing SCGAN’s Disentangled Representation Learning with Contrastive SSIM Similarity Constraints
نویسندگان :
Iman Yazdanpanah
1
Ali Eslamian
2
1- دانشگاه صنعتی اصفهان
2- دانشگاه صنعتی اصفهان
کلمات کلیدی :
Generative Adversarial Nets،Unsupervised Learning،Disentangled Representation Learning،Contrastive Disentanglement،SSIM
چکیده :
SCGAN adds a similarity constraint between generated images and conditions as a regularization term on generative adversarial networks. Similarity constraint works as a tutor to instruct the generator network to comprehend the difference in representations based on conditions. We understand how SCGAN works on a deeper level. This understanding makes us realize that the similarity constraint functions like the contrastive loss function. Two major changes we applied to SCGAN to make a modified model are using SSIM to measure similarity between images and applying contrastive loss principles to the similarity constraint. The modified model performs better using FID and Factor metrics. On the MNIST and Fashion-MNIST datasets, the modified model achieves log-likelihood values of 234.8 and 332.6, respectively, surpassing SCGAN’s 232.5 and 324.2. The modified model exhibits improvements in FID, attaining values of 3.42 and 12.97 for MNIST and Fashion-MNIST, respectively, compared to SCGAN’s 4.11 and 14.63. In terms of disentanglement, the modified model shows clear advantages, achieving values of 0.89 and 0.91 for MNIST and Fashion-MNIST, respectively, compared to SCGAN’s 0.77 and 0.89. These improvements show that the modified model has effectively learned a more disentangled representation compared to SCGAN. The modified model also has better generalisability compared to other models.
لیست مقالات
لیست مقالات بایگانی شده
Energy Efficiency and Pollution Reduction through Intelligent Building Management Systems
Hojjat Hajiabadi - Mahshid Zafari
True Random Number Generator Relying on Multiple Entropy Source and Triple Oscillator for Cryptography Purposes
Somayeh Gholam Mehraban - Mohsen Jalali - Mostafa Azadbakht
اثر پایلوتهای متعامد بر تخمین کانال مایمو انبوه تقسیم فرکانسی مبتنی بر رگرسیون خطی
سید طالب ساداتی لمردی - کمال محامدپور
بهینهسازی مبدلDC-DC منبع امپدانسی مبتنی بر ترانسفورماتور ایزوله گالوانیکی با کاهش تعداد سوئیچها
علی سلیمانی - رضا قاضی
A Transformerless Single-Switch DC-DC Boost Converter Suitable for Renewable Energy Applications
Saed Mahmoud Alilou - Sasan Ahmadi - Mohammad Maalandish - Seyed Hossein Hosseini
Fabrication and performance analysis of a ZnO phototransistor for UV detection
Ghasem Yousefi Simakani - ُSamaneh Hamedi
An Enhanced SLAM Method Using ICP Algorithm for Autonomous Mobile Robots Navigation
Hasan Enami Eraghi - Mohammad Reza Taban - Sayed Farzad Bahreinian - Mohammad Reza Jabbari
طراحی آنتن سرآتش پهن باند مبتنی بر پلاسمون پلاریتونهای سطحی جعلی
فرشاد ارغنده - بیژن عباسی آرند - مریم حصاری شرمه
High-Performance Biosensor Based on SRR for Early Breast Cancer Detection
Hasti Enayattarighehkari - Sina Aramtan - Gholamreza Moradi - Farhad Azadi Namin
تعیین محل خطا با استفاده از اطلاعات حاصل شده از خطا در حضور جبرانساز سری خازنی کنترل تریستوری (TCSC) به روش آفلاین.
حامد حیدری - سعید غنیمتی
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 40.4.2