E-raamat: Image and Graphics: 13th International Conference, ICIG 2025, Xuzhou, China, October 31-November 2, 2025, Proceedings, Part I

Part  1 Artificial Intelligence:


.- SRG-Net: Semantic Relation-Guided Network for Commonsense Video
Captioning.


.- EAANet: Edge-Aware Attention Network for Real-Time Road Scene
Understanding.


.- Learning A Decomposition-Driven Two Stages Unfolding Artifact Removal
Network for Compressed Images.


.- Martingale-Based Skin Lesion Segmentation from Dermoscopic Images.


.- Research on Adaptive Multi-layer Multi-pass Welding Technology for
Medium-Thick Plates.


.- M³E: Mixture of Multi-scale Multi-modal Experts for Time Series
Forecasting.


.- PoseCLR Bridging 2D and 3D Pose Representations via Contrastive Learning
for Action Recognition.


.- Art3D-Fusion: A Hybrid Framework for Visual Synthesis with Artistic
Control.


.- Lesion Localization Prior-Driven Few-Shot Learning for Branch Atheromatous
Disease Diagnosis.


.- Deep Multi-Sentence Aligned Cross-Modal Retrieval.


.- Single-Layer Denoising Taylorformer for UAV Nighttime Tracking.


.- Position-Aware Text-to-Image Generation with Efficient Controllability.


.- Introducing DINOv2 for Medical Image Boundary Tracking.


.- Adaptive Pruning and Cross-Domain Feature Fusion for Robust Object
Tracking.


.- Data Leakage Detection in Large Vision-Language Models via Multimodal
Perturbation.


.- A Novel Dual-Branch Cross-Attention Transformer Network for Low-Dose CT
Denoising.


.- TCGFNet: Multi-Scale Transformer-Convolution with Geometry-Guided Feedback
for Robust Point Cloud Denoising.


.- Adversarial Iterative Pre-Enactment Framework for Air Combat Based on
Mental Simulation Theory.


.- SA-Pillar: Structure-Aware Feature Learning for Real-Time 3D Object
Detection.


.- Knowledge-aware Intent Subgraph Learning for Recommendation.


.- PF-DETR:Enhanced DETR with Pre-Encoded Feature Fusion for Small and
Multi-Scale Object Detection in UAV Imagery.


.- Selective Labeling for 3D Shape Label Transfer based on Local-Global
Features.


.- Part  2 Biological and Medical Image Processing:


.- MAA-Net: A Multi-Attention Aggregation Network for Segmentation of Key
Structures in Microvascular Decompression.


.- Contrastive Hierarchical Graph based Multiple Instance Learning for Fundus
Screening


.- Polyp Segmentation based on Edge Guidance.


.- A Deep Unfolding based on U-Net Graph-Guided Hybrid Regularization  method
for Bioluminescence Tomography.


.- CMambaR: Cardiac Phase Embedded Vision  Mamba for Accelerating Cardiac
MRI  Reconstruction.


.- SC-DSE-nnUNet: An Efficient Hippocampus MRI Segmentation Method.


.- Spatiotemporal Feature Fusion for Glioblastoma Recurrence Prediction Using
Mamba-Based Dual-Stream Framework.


.- Automatic and Fast Segmentation of Cochlear Implant-Induced Artifacts in
MR Images Using Deep Learning.


.- Part  3 Color and Multispectral Processing:


.- End-to-End Diffusion Models with Physics Priors for Enhanced Spectral
Super-Resolution.


.- Asymmetric Dual-Teacher Guided Knowledge Distillation for HSI-SR with
Reconstructed Features.


.- Gradient-based multi-focus image fusion with  focus-aware saliency
enhancement.


.- OME-Net: Optimization-inspired Multi-domain Enhanced Network for Image
Compressed Sensing Reconstruction.


.- Part  4 Compression, Transmission, Retrieval:


.- MARSNet: Scalable Deep Coding of LiDAR Point Clouds via Multimodal and
Residual Learning.


.- Accelerating Learned Video Compression via Low-Resolution Representation
Learning.


.- Optical Flow-driven Fast CU Partition for Inter Prediction in Versatile
Video Coding.


.- Semantic Maintained Video Compression by Background Blurring in
Surveillance Scenarios.


.- Learning Based Fast Coding Unit Decision for Video-based Point Cloud
Compression.


.- Part  5 Computational Imaging:


.- Leveraging a Dual-Learning Methodology Based on Degradation Modeling and
Fractional Fourier Image Transformer for Light Field Image Super-Resolution.


.- Video Stabilization Based on MeshFlow Motion Model in Dynamic and Complex
Scenes.


Dual-Edge Consistency Constrained Unfolding Network for Depth Map
Super-Resolution.


.- Part  6 Computer Graphics and Visualization:


.- Isotropic Remeshing with Inter-Angle Optimization.


.- AlignMR: Design of a Home Yoga Self Learning System Based on MR
Technology.


.- Bi-IRNet: A Transformer-based Binaural Impulse Response Generation
Guidance Model.