Cross-Attention

Follow this topic Ask about this topic

Cross-attention studies how queries from one representation can select and combine relevant information from another, supporting tasks such as image-text understanding, sensor fusion, and sequence forecasting. Common approaches include bidirectional or gated fusion, multi-scale attention, and sparse or geometry-constrained routing, with applications ranging from vision-language models and diffusion generation to medical imaging and robotics. Evaluation spans benchmarks for image recognition, segmentation, multimodal question answering, forecasting, and cross-view matching, with growing emphasis on efficiency and robustness to missing or unreliable inputs. A central tension is that selective attention can improve fusion and control, but attention maps are not always faithful explanations, and simpler concatenation can work better when features are already aligned.

Latest papers

78 papers9 papers in the last four weeks, up 13% on the four weeks before. 0.1% of all new papers.

Geometrical Cross-Attention and Nonvoid Voxelization for Efficient 3D Medical Image Segmentation
Apr 7, 2026Chenxin Yuan, Shoupeng Chen, Haojiang Ye +33D Medical Image Segmentation 3D ViTs
Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, 518100, Guangdong, China · School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, 639798, Singapore · School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen, Shenzhen, 518172, Guangdong, China +2
MTLSI-Net: A Linear Semantic Interaction Network for Parameter-Efficient Multi-Task Dense Prediction
Apr 2, 2026Chen Liu, Hengyu Man, Xiaopeng Fan +1Multi-Task Learning Cross-Attention
Harbin Institute of Technology · Peng Cheng Laboratory · Harbin Institute of Technology SuZhou Research Institute
Hierarchical Pre-Training of Vision Encoders with Large Language Model
Mar 31, 2026Eugene Lee, Ting-Yu Chang, Jui-Huang Tsai +2Vision-Language Pretraining Large Vision-Language Models
University of Cincinnati · National Yang Ming Chiao Tung University
Attention Frequency Modulation: Training-Free Spectral Modulation of Diffusion Cross-Attention
Mar 30, 2026Seunghun Oh, Unsang ParkAttention Control in Diffusion Models Cross-Attention
Sogang University, Seoul, Republic of Korea
EpiMask: Leveraging Epipolar Distance Based Masks in Cross-Attention for Satellite Image Matching
Mar 23, 2026Rahul Deshmukh, Aditya Chauhan, Avinash KakRemote Sensing Image Understanding Image Matching
Purdue University, West Lafayette
Weakly supervised multimodal segmentation of acoustic borehole images with depth-aware cross-attention
Mar 21, 2026Jose Luis Lima de Jesus SilvaMultimodal Semantic Segmentation Image Segmentation
Federal University of Bahia, Institute of Geosciences, Department of Geophysics, Salvador, BA 40170-115, Brazil · Grupo de Estudos e Aplicac¸ ˜ao de Inteligˆencia Artificial em Geof´ısica (GAIA), Federal University of Bahia, Salvador, BA 40170-115, Brazil
Not All Retrievals are Useful: Cross-Attention for Input-Aware RAG in Time Series Forecasting
Mar 16, 2026Seunghan Lee, Jaehoon Lee, Jun Seo +7Zero-Shot Time Series Forecasting Cross-Attention
LG AI Research Seoul, South Korea
DCAU-Net: Differential Cross Attention and Channel-Spatial Feature Fusion for Medical Image Segmentation
Mar 10, 2026Yanxin Li, Hui Wan, Libin LanMedical Image Segmentation Cross-Attention
College of Computer Science and Engineering, Chongqing University of Technology, Chongqing, China
Looking Back and Forth: Cross-Image Attention Calibration and Attentive Preference Learning for Multi-Image Hallucination Mitigation
Mar 7, 2026Xiaochen Yang, Hao Fang, Jiawei Kong +3VLM Hallucination Mitigation Vision-Language Models
Tsinghua Shenzhen International Graduate School, Tsinghua University · Harbin Institute of Technology · Beijing Institute of Technology +1
Making Training-Free Diffusion Segmentors Scale with the Generative Power
Mar 6, 2026Benyuan Meng, Qianqian Xu, Zitai Wang +3Text-Guided Image Segmentation Open-Vocabulary Semantic Segmentation
Institute of Information Engineering, CAS · School of Cyber Security, University of Chinese Academy of Sciences · State Key Laboratory of AI Safety, Institute of Computing Technology, CAS +5
Parallax to Align Them All: An OmniParallax Attention Mechanism for Distributed Multi-View Image Compression
Mar 4, 2026Haotian Zhang, Feiyue Long, Yixin Yu +7Cross-View Attention Learned Image Compression
The National Engineering Laboratory for Video Technology, School of Computer Science, Peking University · Institute for AI Industry Research (AIR), Tsinghua University · Tsinghua Shenzhen International Graduate School, Tsinghua University
Interpretable Cross-Network Attention for Resting-State fMRI Representation Learning
Feb 28, 2026Karanpartap Singh, Adam Turnbull, Mohammad Abbasi +3Brain Network Analysis Functional Connectivity
Stanford University, Stanford, CA 94305
TG-ASR: Translation-Guided Learning with Parallel Gated Cross Attention for Low-Resource Automatic Speech Recognition
Feb 25, 2026Cheng-Yeh Yang, Chien-Chun Wang, Li-Wei Chen +3Automatic Speech Recognition Speech Processing
Dept. Computer Science and Information Engineering, National Taiwan Normal University, Taiwan · United Link Co., Ltd., Taiwan · Institute of Computer Science, Academia Sinica, Taiwan
Effective and Robust Multimodal Medical Image Analysis
Feb 17, 2026Joy Dhar, Nayyar Zaidi, Maryam HaghighatMedical Image Analysis Multimodal Fusion
Indian Institute of Technology Ropar, Rupnagar, Punjab, India · Deakin University, Melbourne, Victoria, Australia · Queensland University of Technology, Brisbane, QLD, Australia
Scalable Cross-Attention Transformer for Cooperative Multi-AP OFDM Uplink Reception
Feb 4, 2026Xavier Tardy, Grégoire Lefebvre, Apostolos Kountouris +2Transformer Cross-Attention
IETR - UMR CNRS 6164, CentraleSup´elec, avenue de la Boulaie - CS 47601 35576 CESSON-SEVIGNE Cedex, France · Orange Research, Grenoble, France
AttentionViG: Cross-Attention-Based Dynamic Neighbor Aggregation in Vision GNNs
Sep 29, 2025Hakan Emre Gedik, Andrew Martin, Mustafa Munir +4Graph Attention Networks Cross-Attention
The University of Texas at Austin
Cross-Attention is Half Explanation in Speech-to-Text Models
Sep 22, 2025Sara Papi, Dennis Fucci, Marco Gaido +2Transformer Interpretability Cross-Attention
Fondazione Bruno Kessler, Italy
CrossFusion: A Multi-Scale Cross-Attention Convolutional Fusion Model for Cancer Survival Prediction
Mar 3, 2025Rustin Soraki, Huayu Wang, Sitong Liu +2Computational Pathology Cancer Survival Prediction
University of Washington, Seattle, WA · University of California, Los Angeles, CA

Apr 7, 2026J/K move · Enter open · S save

Geometrical Cross-Attention and Nonvoid Voxelization for Efficient 3D Medical Image Segmentation

Chenxin Yuan, Shoupeng Chen, Haojiang Ye, Yiming Miao+2

Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, 518100, Guangdong, China · School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore, 639798, Singapore · School of Science and Engineering, The Chinese University of Hong Kong, Shenzhen, Shenzhen, 518172, Guangdong, China+2

3D Medical Image Segmentation 3D ViTs Cross-Attention

Open paper AskPDF arXiv

Accurate segmentation of 3D medical scans is crucial for clinical diagnostics and treatment planning, yet existing methods often fail to achieve both high accuracy and computational efficiency across diverse anatomies and imaging modalities. To address these challenges, we propose GCNV-Net, a novel 3D medical segmentation framework that integrates a Tri-directional Dynamic Nonvoid Voxel Transformer (3DNVT), a Geometrical Cross-Attention module (GCA), and Nonvoid Voxelization. The 3DNVT dynamically partitions relevant voxels along the three orthogonal anatomical planes, namely the transverse, sagittal, and coronal planes, enabling effective modeling of complex 3D spatial dependencies. The GCA mechanism explicitly incorporates geometric positional information during multi-scale feature fusion, significantly enhancing fine-grained anatomical segmentation accuracy. Meanwhile, Nonvoid Voxelization processes only informative regions, greatly reducing redundant computation without compromising segmentation quality, and achieves a 56.13% reduction in FLOPs and a 68.49% reduction in inference latency compared to conventional voxelization. We evaluate GCNV-Net on multiple widely used benchmarks: BraTS2021, ACDC, MSD Prostate, MSD Pancreas, and AMOS2022. Our method achieves state-of-the-art segmentation performance across all datasets, outperforming the best existing methods by 0.65% on Dice, 0.63% on IoU, 1% on NSD, and relatively 14.5% on HD95. All results demonstrate that GCNV-Net effectively balances accuracy and efficiency, and its robustness across diverse organs, disease conditions, and imaging modalities highlights strong potential for clinical deployment.