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

.- Part  7 Computer Vision and Pattern Recognition:


.- Unsupervised image restoration using domain discriminator with feature
disentangle.


.- Cross-domain Adaptation for Few-shot 3D Shape Generation.


.- SaliencyCLIP-SAM: Bridging Text and Image Towards Text-driven Salient
Object Detection.


.- Efficient RGBT Tracking via Early Fusion and Hierarchical Knowledge
Distillation.


.- Human Pose Estimation Method Based on Top-down View Fisheye Images.


.- On Leveraging Unlabeled Data for Concurrent Positive-Unlabeled
Classification and Robust Generation.


.- Multi-View Captioning with Semantic Delta Re-Ranking for Zero-Shot
Composed Video Retrieval.


.- Feature Decoupling with Modality Modulation for Multimodal Sentiment
Analysis.


.- A Two-stage Multimodal Remote Sensing Image Registration Network with
Deformation-Refined Affine Transformation.


.- FFTA-Net A Frequency-Domain Fusion and Temporal Alignment Network for
Transmission Line Defect Detection.


.- Hierarchical Transformer for Panoramic Image Inpainting with Comprehensive
Attention Module.


.- Time-Frequency Domain-based No-reference Algorithm for Image Blurriness
Evaluation.


.- Boosting Portrait Matting with Spatial-Frequency Harmony.


.- w+:Extending Classifier-Free Guidance in Diffusion Models for Real Image
Inversion.


.- Self-Supervised Surface Defect Inspection Method via Alignment of Content
and Style.


.- Improved UAV aerial vehicle detection algorithm based on YOLOv11n.


.- NSFM: Normalization-based Style Feature Elimination Method for Remote
Sensing Scene Classification.


.- Hyperspectral Image Super-resolution via Degradation-aware Learning and
Frequency-domain Feature Enhancement.


.- Nighttime Object Detection with Contextual Auxiliary Learning.


.- Learning Supplementary Information for  First-Person Perception Referring
Expression  Compression.


.- MFMamba: Multiple Fusing Mamba Network for Hyperspectral Image
Pansharpening.


.- Point Density Fusion for Multimodal 3D Object  Detection.


.- Temporal Reordering for Video Person Re-Identification Based on Feature
Reappearance Score.


.- Multimodal Consistency-Driven Deepfake Detection.


.- Leveraging Large-Scale Pretrained Vision Foundation Models for
Label-Efficient 3D Point Cloud Segmentation.


.- Video Domain Incremental Learning for Human Action Recognition in Home
Environments.


.- Wavelet-Based Distillation with Structured Frequency Alignment.


.- A Vessel Extraction Method Based on Bilinear Factor Matrix Norm RPCA and
TSRG.


.- A Progressive Approach to Learn Global and Local Multi-View Features for
3D Visual Grounding.


.- HG-Ghost: A Lightweight and Accurate Pose Estimation Network for Biometric
Recognition.


.- AMFlow: Efficient Optical Flow Estimation via Attentional Cost Volume and
Matching Initialization.


.- Fighting Detection Based on Individual Keypoints Motion Trajectories and
Motion Direction Entropies.


.- Design and experiment of online non-destructive quality testing and
grading device for chicory sprouts.


.- Exploring Implicit Relations for Fine-grained Generalized Category
Discovery.


.- SPViT-FER: A Sparse Pruning Based Vision Transformer for Facial Expression
Recognition.


.- Deciphering the Visual Style of Chinas Hit Short Videos through Computer
Vision.


.- A lightweight and real-time asymmetric multi-output thermal radiation
effects correction in infrared images.


.- SNN-PAR: Energy Efficient Pedestrian Attribute Recognition via Spiking
Neural Networks.


.- GramFormer-based Crowd Counting with Learnable Fourier Encoding and
Attention Mechanism.


.- TEI-Face: A Temporal Expression and Identity Stability Oriented Face
Swapping.


.- COS-SLAM: Coordinate Attention Semantic SLAM with Pixel-to-Line
Transformer.


.- Perspective Driven Prototype Alignment for Aerial-Ground Person
Re-identification.


.- Robust Single-View 3D Object Reconstruction with Stable Diffusion
Generation and Farthest View Selection.


.- From Sky to Site: A Unified Framework for Static and Dynamic 3D
Reconstruction in .- Construction Sites.


.- Adaptive Positional Encoding and Multi-scale Self-attention Transformer
for Aerial Person Re-Identification.


.- Class-incremental learning for surface defect detection.


.- Video Frame Interpolation via Iterative Optical Flow Refinement with
Latent Motion Feature.