Biometric Recognition: 19th Chinese Conference, CCBR 2025, Nanchang, China, November 21–23, 2025, Proceedings(16360 Lecture Notes in Computer Science)

LNCS 16360 constitutes the proceedings of the 19th Chinese Conference on Biometric Recognition, CCBR 2025, held in Nanchang, China, during November 21–23, 2025. The 62 full papers presented here were carefully reviewed and selected from 90 submissions. They were organized in the following topical sections: Fingerprint, Palmprint and Vein Recognition; Human-Centric AIGC (Face Synthesis, Speech Synthesis, Gesture Generation, Human Motion Generation, etc.); Gait, Footprint;  Face Related; Emotional, Psychological, Physiological, and Health Intelligence Perception and Computing; Biometrics in Mobile Terminals, Healthcare, Banking, Internet of Things, Self-Driving,  Intelligent Robots, etc; Anti-spoofing, Presentation Attack Detection; Human-Related Understanding; Basic Theory of Biometric Recognition; Gesture, Action; Individual Characterization and Human-Computer Interaction; Adversarial Attack and Proactive Defense; Adversarial Attack and Proactive Defense; Template Protection and Cryptosystems; Datasets, Evaluation, Benchmarking, Performance Modelling and Prediction; Others.

Table of Contents:
.- Fingerprint, Palmprint and Vein Recognition. .- EC-PVGAN : Affine-Equivariant Generative Adversarial Data Augmentation for Palm-Vein Identification. .- A cloud-edge-end federated learning secret sharing scheme for finger vein recognition system. .- Toward Abandoning Tedious ROI Alignment for Unconstrained Palmprint Recognition. .-  Contact-to-Contactless Fingerprint Generation with Content Consistency. .- UMR-Net: Unified Multimodal Representation Network for Multimodal Biometric Recognition with Missing Modality. .- Enhanced Contactless Palmprint Backdoor Attack with Invisible Sample-Specific Triggers. .-  Progressive Adversarial Learning for Multi-Modality Biometric Recognition with Missing Modality. .- Learning Without Borders: A Domain-Adapted and Federated Approach to Palmprint Recognition. .- Cross-Domain Palmprint Cryptosystems via Image Alignment and Neural Error Correction. .- TSCAN: Teacher-Student Co-Learning Adaptive Network for Cross-Device Palmprint Recognition. .- Internal Fingerprint Imaging System Based On Full Field Optical Coherence Tomography. .- AReview on Palmprint Image-level Attacks. .- Topographic Feature-Based Vein Biometric Recognition. .- RSANet Multi-level Fusion Dual-modal Recognition Network. .- Dynamic Selective Distillation Network Based on Quality-Aware Fusion for Multimodal Biometric Recognition LACE: LearnableAdaptive Cross-Entropy Loss for vein recognition  .- Human-Centric AIGC (Face Synthesis, Speech Synthesis, Gesture Generation, Human Motion Generation, etc.). .- EmoPrompt+: Emotional Image Content Generation via Emotion-Driven Prompting and Multi-Level Emotional Guidance in Stable Diffusion. .- Online Emotion-Driven Generation of Multiple Appropriate Facial Reactions. .- HairEditor: Diffusion-Guided Supervision for StyleGAN-Based Hair Editing in Real-World Portraits. .- Conservation-informed Neural Network for Human Motion Prediction. .- High-Low Feature Fusion Generative Adversarial Network for the Inpainting of Irregularly Occluded Iris Images. .- FP-Director: Direction-Guided Latent Code Refinement for Facial-Preference Alignment in Text-to-Image Diffusion. .- Hand Motion Retargeting Based on GraphAttention Residual Perception. .- Gait, Footprint. .- SMEGNet:ALightweight MLP-EnhancedArchitecture for Cross-View Gait Recognition. .- Revisiting Euclidean Triplet Loss for Gait Recognition. .- Action-agnostic Pose-based Gait Recognition. .- DeepSNNGait:ASpiking Neural Network Framework for Robust Gait Recognition. .- HealthGait-Uni: Health Assessment by Human Body Appearance and Motion from  Videos. .- Face Related. .- DRAge: Dynamic Routing Mixture of Experts for Facial Age Estimation. .- PGS-Net: Personalized Graph Structure Network with Self-Supervised Learning for Micro-Expression Recognition. .- Cross-Paradigm Facial Expression Recognition based Emotional Category-Feature Prototypes. .- Attribute-Driven Identity Disentanglement for Fine-Grained Face Anonymization. .-  Emotional, Psychological, Physiological, and Health Intelligence Perception and Computing. .- AU-LLM: Micro-ExpressionAction Unit Detection via Enhanced LLM-Based Feature Fusion.  .- TwoM:An EEG depression identification model via spatiotemporal filtering. .- Learning to Decompose and Fuse A Hybrid Approach for Noise-Robust Remote Photoplethysmography. .- Automatic Sleep Stage Classification with Hypergraph Neural Networks Using Spatial-Temporal Features. .- Biometrics in Mobile Terminals, Healthcare, Banking, Internet of Things, Self-Driving, Intelligent Robots, etc. .- Research on DNA storage encoding methods and evaluation standard system .- MSPD-SAM: A Prompt-Free Framework for Cardiac Segmentation using Multi-Scale Adapters and Parallel Decoding. .- MGONet:An Optimized Segmentation Network for Esophageal Cancerous Lesions. .- Automatic Assessment of Facial Paralysis Severity from 3D Point Clouds. .- AM-UNet:Attention Mamba U-Net for Medical Image Segmentation. .- Multi-scale Channel Attention Vision LSTM Network for Optic Cup and Optic Disc Segmentation. .- Anti-spoofing, Presentation Attack Detection. .- Deep Learning-Based Approaches for Iris Image Spoofing Prevention and Tamper Detection. .- Fingerprint Liveness Detection Based on EfficientNet andAdversarial attacks. .- Bridging Synthetic and Real Domains for Face Presentation Attack Detection via Entropy-Regularized Alignment. .-  Adapting Vision Transformer with Dual Stream Token Difference for Mobile Face Anti-Spoofing. .- Domain Generalization in Face Anti-Spoofing based on Vision-Language Semantic Awareness. .- Human-Related Understanding. .- Distribution-discriminative and Modality-aware Test-time Cross-domain Adaptation for Text-based Person Search. .- MFNet: Mamba-Driven Feature Fusion for Human Parsing. .- Automatic Visual-Language Aligning Network for Visible-Infrared Person Re-Identification. .- Temporally-Aware Multi-task Representation Learning for Compositional Action Recognition. .- CGDRF-YOLO:Alightweight and efficient UAV-based pedestrian detection algorithm. .- Region-level Cross-modal Matching Framework for Text-based Geo-localization. .- Basic Theory of Biometric Recognition. .- Exponential Non-negative Matrix Factorization for Image Data Representation. .- Orthogonal-Bidirectional Pose Anchoring Model for Micro-Expression Recognition. .- Gesture, Action. .-  A Unified Transformer with a Parametric Activation Function for Robust Gesture Recognition across Sparse and Dense EMG Signals. .- Multimodal Higher-Order Statistical Adapter For Video Action Recognition. .- Individual Characterization and Human-Computer Interaction. .- Context- and Visibility-Aware Part Learning for Aerial-Ground Person Re-Identification. .- Adversarial Attack and Proactive Defense. .-  Adversarial Prompt Increment for Robust Vision-Language Models. .- Template Protection and Cryptosystems. .- A highly secure biometric template protection method based on Householder matrices and Absolute Value Equation Transform. .- Datasets, Evaluation, Benchmarking, Performance Modelling and Prediction. .- Study on the Construction of a Biometric Database of Parasites in Cattle and Sheep. .- Others. .- Disentangled Representation Learning for Single-Domain Generalization in PPG Biometric Recognition.