AIConfPaper
← Search

Ulas Bagci

11 accepted papers

2026

ALIGNING WHAT YOU SEPARATE: DENOISED PATCH MIXING FOR SOURCE-FREE DOMAIN ADAPTATION IN MEDICAL IMAGE SEGMENTATION

Quang-Khai Bui-Tran, Thanh-Huy Nguyen, Hoang-Thien Nguyen, Ba-Thinh Lam, Nguyen Lan Vi Vu, Phat Huynh +2
ICASSP 2026poster

Source-Free Domain Adaptation (SFDA) is emerging as a compelling solution for medical image segmentation under privacy constraints, yet current approaches often ignore sample difficulty and struggle with noisy supervision under domain shift. We present a new SFDA framework that leverages Hard Sample…

Cited by 0Source ↗PDF ↗Scholar ↗
2026

DOMAIN-INVARIANT MIXED-DOMAIN SEMI-SUPERVISED MEDICAL IMAGE SEGMENTATION WITH CLUSTERED MAXIMUM MEAN DISCREPANCY ALIGNMENT

Ba-Thinh Lam, Thanh-Huy Nguyen, Hoang-Thien Nguyen, Nguyen Lan Vi Vu, Quang-Khai Bui-Tran, Ulas Bagci +1
ICASSP 2026poster

Deep learning has shown remarkable progress in medical image semantic segmentation, yet its success heavily depends on large-scale expert annotations and consistent data distributions. In practice, annotations are scarce, and images are collected from multiple scanners or centers, leading to mixed-d…

Cited by 0Source ↗PDF ↗Scholar ↗
2026

Discrete Cosine Transform Based Decorrelated Attention for Vision Transformers

Hongyi Pan, Emadeldeen Hamdan, Xin Zhu, Ahmet Enis Cetin, Ulas Bagci
IJCAI 2026

Self-attention is central to the success of Transformer architectures; however, learning the query, key, and value projections from random initialization remains challenging and computationally expensive. In this paper, we propose two complementary methods that leverage the Discrete Cosine Transform

Cited by 0Scholar ↗
2026

LUMINA: A Multi-Vendor Mammography Benchmark with Energy Harmonization Protocol

Hongyi Pan, Gorkem Durak, Halil Ertugrul Aktas, Andrea M. Bejar, Baver Tutun, Emre Uysal +6
CVPR 2026

Publicly available full-field digital mammography (FFDM) datasets remain limited in size, clinical annotations, and vendor diversity, hindering the development of robust models. We introduce LUMINA, a curated, multi-vendor FFDM dataset that explicitly encodes acquisition energy and vendor metadata t

Cited by 0Source ↗code ↗Scholar ↗
2025

Frequency-Based Federated Domain Generalization for Polyp Segmentation

Hongyi Pan, Debesh Jha, Koushik Biswas, Ulas Bagci
ICASSP 2025accepted

Federated Learning (FL) offers a powerful strategy for training machine learning models across decentralized datasets while maintaining data privacy, yet domain shifts among clients can degrade performance, particularly in medical imaging tasks like polyp segmentation. This paper introduces a novel…

Cited by 0Source ↗Scholar ↗
2025

MDNet: Multi-Decoder Network for Abdominal CT Organs Segmentation

Debesh Jha, Nikhil Kumar Tomar, Koushik Biswas, Gorkem Durak, Matthew Antalek, Zheyuan Zhang +10
ICASSP 2025accepted

Accurate segmentation of organs from abdominal CT scans is essential for clinical applications such as diagnosis, treatment planning, and patient monitoring. To handle challenges of heterogeneity in organ shapes, sizes, and complex anatomical relationships, we propose a Multi decoder network (MDNet)…

Cited by 0Source ↗Scholar ↗
2025

Order-aware Interactive Segmentation

Bin Wang, Anwesa Choudhuri, Meng Zheng, Zhongpai Gao, Benjamin Planche, Andong Deng +4
ICLR 2025poster

Interactive segmentation aims to accurately segment target objects with minimal user interactions. However, current methods often fail to accurately separate target objects from the background, due to a limited understanding of order, the relative depth between objects in a scene. To address this is…

Cited by 0Source ↗PDF ↗Scholar ↗
2025

Transformer-Enhanced Iterative Feedback Mechanism For Polyp Segmentation

Nikhil Kumar Tomar, Debesh Jha, Koushik Biswas, Ulas Bagci
ICASSP 2025accepted

Colorectal cancer (CRC) is the third most common cause of cancer diagnosed in the United States. Notably, CRC is the leading cause of cancer in younger men less than 50 years old. Colonoscopy is considered the gold standard for the early diagnosis of CRC. Skills vary significantly among endoscopists…

Cited by 0Source ↗Scholar ↗
2025

ViCTr: Vital Consistency Transfer for Pathology Aware Image Synthesis

Onkar Susladkar, Gayatri Deshmukh, Yalcin Tur, Gorkem Durak, Ulas Bagci
ICCV 2025poster

We introduce ViCTr (Vital Consistency Transfer), a framework for advancing medical image synthesis through a principled integration with Rectified Flow trajectories. Unlike traditional approaches, we modify the Tweedie formulation to accommodate linear trajectories within the Rectified Flow framewor…

Cited by 0Source ↗PDF ↗code ↗Scholar ↗
2025

VideoAds for Fast-Paced Video Understanding

Zheyuan Zhang, Wanying Dou, Linkai Peng, Hongyi Pan, Ulas Bagci, Boqing Gong
ICCV 2025accepted

Advertisement videos serve as a rich and valuable source of purpose-driven information, encompassing high-quality visual, textual, and contextual cues designed to engage viewers. They are often more complex than general videos of similar duration due to their structured narratives and rapid scene tr…

Cited by 0Source ↗PDF ↗Scholar ↗
2024

Domain Generalization with fourier Transform and soft thresholding

Hongyi Pan, Bin Wang, Zheyuan Zhang, Xin Zhu, Debesh Jha, Ahmet Enis Çetin +2
ICASSP 2024accepted

Domain generalization aims to train models on multiple source domains so that they can generalize well to unseen target domains. Among many domain generalization methods, Fourier-transformbased domain generalization methods have gained popularity primarily because they exploit the power of Fourier t…

Cited by 0Source ↗Scholar ↗