Erkut Erdem's Avatar

Erkut Erdem

@erkuterdem.bsky.social

Professor of Computer Science at Hacettepe University, Researcher in Computer Vision, Machine Learning, Artificial Intelligence, TUBA-GEBIP Outstanding Young Scientist Awardee

293 Followers  |  182 Following  |  4 Posts  |  Joined: 23.02.2023
Posts Following

Posts by Erkut Erdem (@erkuterdem.bsky.social)

A huge thanks to the incredible team behind HyperGAN-CLIP! Abdul Basit Anees, Ahmet Canberk Baykal, Muhammed Burak Kizil, Duygu Ceylan @aykuterdem.bsky.social ๐Ÿ™Œ (4/n)

21.11.2024 05:51 โ€” ๐Ÿ‘ 2    ๐Ÿ” 0    ๐Ÿ’ฌ 0    ๐Ÿ“Œ 0
Post image

๐ŸฎCurious to learn more? Catch us at #SIGGRAPHAsia 2024 in Tokyo or check out our paper for all the details.

Project page: cyberiada.github.io/HyperGAN-CLIP
Arxiv: arxiv.org/abs/2411.12832
(3/n)

21.11.2024 05:51 โ€” ๐Ÿ‘ 2    ๐Ÿ” 1    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
Post image

Our approach extends a pre-trained StyleGAN by integrating CLIP space via hypernetworks. This allows us to dynamically adapt it to new domains using reference images or text descriptions.

Itโ€™s flexible, efficient, and unlocks new applications for GANs. ๐ŸŒโœจ (2/n)

21.11.2024 05:51 โ€” ๐Ÿ‘ 2    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 0
A Unified Framework for Domain Adaptation, Image Synthesis and Manipulation (SIGGRAPH Asia'24)
YouTube video by Erkut Erdem A Unified Framework for Domain Adaptation, Image Synthesis and Manipulation (SIGGRAPH Asia'24)

GANs like StyleGAN generate highly realistic images. But adapting them to new domains or tasks like text-guided editing or reference-guided synthesis with limited data is challenging! ๐Ÿ–ผ๏ธโœจ

Our #SIGGRAPHAsia 2024 paper, HyperGAN-CLIP, tackles this: youtu.be/X0VOYFhPWxQ (1/n)

21.11.2024 05:51 โ€” ๐Ÿ‘ 2    ๐Ÿ” 0    ๐Ÿ’ฌ 1    ๐Ÿ“Œ 2