🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

This week we will host Assoc. Prof. Aziz F. Zambak from Department of Philosophy at METU to talk about the main ontological commitments of artificial intelligence. ⏰24 Dec, Friday, 12:40 (Turkish Time) 🌐 zoom.us/j/91741847015?…

all you need is a good idea:

New #WACV2022 paper: "MaskSplit: Self-supervised Meta-learning for Few-shot Semantic Segmentation", by Sercan Amaç, Ahmet Şencan, Bugra Baran, Nazli Ikizler Cinbis, and Gokberk Cinbis. Joint work with HUCVL. Paper: openaccess.thecvf.com/content/WACV20… Code: github.com/AhmetSencan/Ma…

👉👉Upcoming Seminar! The CEO and co-founder of Locomation Locomation, Inc., Dr. Çetin Meriçli Çetin Meriçli will be with us on 14 January, at 16:30 (Turkish Time). 🌐Join us zoom: zoom.us/j/91741847015?… Meeting ID: 917 4184 7015 Passcode: 2022

#ODTÜMezuniyetininYeriDevrimdir

#ODTÜMezuniyetininYeriDevrimdir

CERN’de, ODTÜ bilgisayar bölümünden Emre Akbaş hocayla verilerin makine öğrenmesi konusundaki işbirliğimizi güçlendirdik, NASA’dan Kathy Lueders’i ağırladık ve ODTÜ ekibinden 3 öğrencimle güzel işler yaptık. Artık yurda dönme vakti …

That's exactly what freshman researchers are looking for :)

At 2100+ (not a typo) pages 🤯, this is almost all the Math you need for Machine Learning without dumbing it down! PDF link: cis.upenn.edu/~jean/math-dee…

haha xd

Diffusion models for handwriting generation! Cool work. arxiv.org/pdf/2011.06704…

🤖📣We will be hosting Erdem Biyik of UC Berkeley on Monday evening! Join us on Zoom. As usual more info and links available at: romer.metu.edu.tr/talks Erdem Bıyık

Excited about our IJCAI2023 paper on fairness in mental health prediction. Our paper has counterintuitive results about worse predictions on certain demographic groups, despite having more date. We show that balanced/more data does not mean fairer/better predictions in general.

Differentiable max & argmax operators are defined through concave maximization on the simplex. Soft and sparse max are particular cases. arxiv.org/abs/1602.02068 arxiv.org/abs/1705.07704 arxiv.org/abs/1802.03676 arxiv.org/abs/1802.04223

New video series: Physics Informed Machine Learning! Physics may be embedded into AI/ML in 5 stages: 1 choose what to Model 2 curate training Data 3 design an Architecture 4 craft a Loss Function, and 5 implement Optimization Algorithm to train the model youtube.com/watch?v=JoFW2u…

"RankED: Addressing imbalance and uncertainty in edge detection using ranking based losses" being presented at #cvpr2024 right now. Actual reviewer comments in red.

Happy to announce that our work (with Barış Can, Adnan Harun DOĞAN, Kemal Öksüz, Emre Akbas and Sinan Kalkan) titled ‘Bucketed Ranking-based Losses for Efficient Training of Object Detectors’ has been accepted to #ECCV2024! [1/2]

Barış Can Adnan Harun DOĞAN Kemal Öksüz Emre Akbas Sinan Kalkan We show in our paper that we can significantly (up to 6x) improve the efficiency of ranking-based losses and use them to train a DETR-based detector for the first time. Paper, code and details to follow soon. [2/2]

Check out our poster (#39) between 10.30-12.30 today! #ECCV2024

Feyza is presenting our poster now at #ECCV2024. We introduce Bucketed Ranking-based Losses to directly optimize performance measures of object detectors in an efficient manner and use them to improve various detectors. Barış Can Adnan Harun DOĞAN Emre Akbas Sinan Kalkan