Category Ranking
98%
Total Visits
921
Avg Visit Duration
2 minutes
Citations
20
Article Abstract
Domain Adaptive Object Detection (DAOD) focuses on improving the generalization ability of cross-domain object detection via knowledge transfer. Recent advances in DAOD strive to minimize the domain discrepancy for adapting class-conditional distributions by aligning cross-domain sampled node pairs in the non-Euclidean graphical space. However, these methods of reformulating the adaptation with graph matching may fail to concentrate on precise semantics alignment, due to the domain-biased graph modeling and the unreliable matching with background nodes. To solve these issues, a Foreground-aware Universe Graph Matching (FUGM) framework is proposed for DAOD. Specifically, a virtual universe graph is category-wise constructed for modeling semantic knowledge, and node representation is refined via Collaborative Graph Reasoning (CGR), which incorporates with self-loop calibration for the information that node learns about itself and universe graph based semantic interaction. To reduce the spurious matching pair for better semantics alignment, Universe Graph Matching (UGM) is developed to encourage instance nodes to match anchor nodes and remove absorbing nodes. The resulting matching model enables end-to-end learning especially for instance node pairwise affinity enhancement through gathering foreground nodes of corresponding category together. Extensive experiments verify that FUGM outperforms existing works significantly.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1016/j.neunet.2025.107804 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Association between maternal circulating total triiodothyronine and subsequent risk of preeclampsia.
Eur J Med Res
September 2025
Obstetrics and Gynecology Hospital, Fudan University, 566 Fangxie Road, Shanghai, 200011, China.
Introduction: To clarify the uncertain association between maternal total triiodothyronine (TT3) levels and preeclampsia risk.
Methods: In a hospital-based cohort of pregnant women with universal thyroid testing, we assessed the association between TT3 and preeclampsia using directed acyclic graphs (DAG) to control confounders.
Results: Maternal TT3 levels were associated with preeclampsia risk, with an adjusted odd ratio (OR) of 2.
Towards Generic Abdominal Multi-Organ Segmentation with multiple partially labeled datasets.
Comput Med Imaging Graph
September 2025
Shanghai Key Laboratory of Trustworthy Computing, East China Normal University, Shanghai 200062, China.
An increasing number of publicly available datasets have facilitated the exploration of building universal medical segmentation models. Existing approaches address partially labeled problem of each dataset by harmonizing labels across datasets and independently focusing on the labeled foreground regions. However, significant challenges persist, particularly in the form of cross-site domain shifts and the limited utilization of partially labeled datasets.
View Article and Find Full Text PDFFoldExplorer: Fast and Accurate Protein Structure Search with Sequence-Enhanced Graph Embedding.
J Mol Biol
August 2025
Institute of Image Processing and Pattern Recognition, Shanghai Jiao Tong University, and Key Laboratory of System Control and Information Processing, Ministry of Education of China, Shanghai 200240, China. Electronic address:
The advent of highly accurate protein structure prediction methods has fueled an exponential expansion of the protein structure database. Consequently, there is a rising demand for rapid and precise protein structure search. Traditional alignment-based methods are designed for precise pairwise comparisons, offering high accuracy.
View Article and Find Full Text PDFSignificantly enhancing human antibody affinity via deep learning and computational biology-guided single-point mutations.
Brief Bioinform
August 2025
Faculty of Pharmaceutical Sciences, Shenzhen University of Advanced Technology, 1 Beizhen Road, Xinhu Subdistrict, Guangming District, Shenzhen 518055, China.
Enhancing antibody affinity is a critical goal in antibody design, as it improves therapeutic efficacy, specificity, and safety while reducing dosage requirements. Traditional methods, such as single-point mutations or combinatorial mutagenesis, are limited by the impracticality of exhaustively exploring the vast mutational space. To address this challenge, we developed a novel computational pipeline that integrates evolutionary constraints, antibody-antigen-specific statistical potentials, molecular dynamics simulations, metadynamics, and a suite of deep learning models to identify affinity-enhancing mutations.
View Article and Find Full Text PDFA Universal Machine Learning Framework Driven by Artificial Intelligence for Ion Battery Cathode Material Design.
JACS Au
August 2025
College of Chemistry, Key Laboratory of Theoretical & Computational Photochemistry of Ministry of Education, Beijing Normal University, Beijing 100875, People's Republic of China.
Graph neural networks for crystal property prediction typically require precise atomic positions and types, limiting their applicability for novel materials with unknown structures. To address this limitation, we introduce BatteryFormer, a versatile machine learning model that employs average interatomic radius distance instead of precise bond lengths as edge embedding, enabling rapid, high-throughput material screening based solely on composition and structural prototypes. BatteryFormer demonstrates robust predictive performance across a wide range of intervals.
View Article and Find Full Text PDF