Emre Akbas

Emre Akbas is an Associate Professor at the Department of Computer Engineering, Middle East Technical University (METU). He previously worked as a researcher at the Vision and Image Understanding Lab at University of California Santa Barbara (UCSB) and earned his PhD from University of Illinois at Urbana-Champaign (UIUC) under Prof. Narendra Ahuja. His academic journey includes MS and BS degrees from METU, where he ranked 1st in his undergraduate class.

• Education: PhD (UIUC), MS (METU), BS (METU)
• Honors: Young Scientist Award (2022), METU Thesis Awards (2023, 2022, 2020), Beckman Institute Cognitive Science/AI Award (2010)

His research focuses on visual detection and machine learning, particularly addressing imbalance problems in object detection. He co-developed innovative methods like HoughNet, a generalized Hough transform for object detection, and formulated the Localization Recall Precision (LRP) metric for visual detection evaluation. His work spans deep learning, domain adaptation, and generative adversarial networks (GANs).

Recent publications highlight his contributions to object detection metrics, ranking-based loss functions, and domain adaptation techniques. His research group has produced significant works at venues like CVPR, ECCV, and TPAMI, including the textbook Signals and Systems: Theory and Practical Explorations with Python (Wiley, 2024).

• Awards: Young Scientist Award (Science Academy, Turkey), METU Thesis Awards
• Grants: ERC Starting Grant (NONWESTLIT), TUBITAK Career Development Grant, AWS Cloud Credits

As an advisor, he has guided multiple thesis awardees. His community service includes organizing ECCV and CVPR workshops and serving as a reviewer for top-tier conferences and journals.