Cross-Camera Deep Colorization -- Attentive Cascaded Pyramid Network for Online Video Stabilization -- Amodal Layout Completion in Complex Outdoor Scenes -- Exploring Hierarchical Prototypes for Few-Shot Segmentation -- BSAM: Bidirectional Scene-Aware Mixup for Unsupervised Domain Adaptation in Semantic Segmentation -- Triple GNN: A Pedestrian-Scene-Object Joint Model for Pedestrian Trajectory Prediction -- Cross-Domain Trajectory Prediction With CTP-Net -- Spatial-Aware GAN for Instance-Guided Cross-Spectral Face Hallucination -- Lightweight Image Compression Based on Deep Learning -- DGMLP: Deformable Gating MLP Sharing for Multi-Task Learning -- Monocular 3D Face Reconstruction with Joint 2D and 3D Constraints -- Scene Text Recognition with Single-Point Decoding Network -- Unsupervised Domain Adaptation for Semantic Segmentation with Global and Local Consistency -- Research on Multi-Temporal Cloud Removal using D-S Evidence Theory and Cloud Segmentation Model -- SASD: A Shape-Aware Saliency Object Detection Approach for RGB-D Images -- Dual Windows Are Significant: Learning from Mediastinal Window and Focusing on Lung Window -- CDNeRF: A Multi-Modal Feature Guided Neural Radiance Fields -- MHPro: Multi-Hypothesis Probabilistic Modeling for Human Mesh Recovery -- Image Sampling for Machine Vision -- SMOF: Squeezing More Out of Filters Yields Hardware-Friendly CNN Pruning -- H-ViT: Hybrid Vision Transformer for Multi-Modal Vehicle Re-Identification -- A Coarse-to-Fine Convolutional Neural Network for Light Field Angular Super-Resolution -- VSA: Adaptive Visual and Semantic Guided Attention on Few-Shot Learning -- Evolutionary Multitasking for Coarse-to-Fine Point Cloud Registration with Chaotic Opposition Search Strategy -- Matching and Localizing: A Simple Yet Effective Framework for Human-Centric Spatio-Temporal Video Grounding -- Part-Wise Topology Graph Convolutional Network for Skeleton-Based Action Recognition -- A Lightweight Detection Method of Smartphone Assembly Parts -- Illumination-Guided Transformer-Based Network for Multispectral Pedestrian Detection -- 3D Face Cartoonizer: Generating Personalized 3D Cartoon Faces from 2D Real Photos with a Hybrid Dataset -- Low Light Video Enhancement Based on Temporal-Spatial Complementary Feature -- Compound Label Learning for Affective Image Content Analysis -- Weakly Supervised Learning of Instance Segmentation with Confidence Feedback -- Robust Face-Swap Detection Based on 3D Facial Shape Information -- Prompt Learning with Cross-Modal Feature Alignment for Visual Domain Adaptation -- Tile Defect Detection Based on the Improved S2C-YOLOv5 -- Human-Object Interaction Detection: A Survey of Deep Learning-based Methods -- Change Detection of Flood Hazard Areas in Multi-Source Heterogeneous Earth Observation Image Time Series Based on Spatiotemporal Enhancement Strategy -- Reinforcement Learning Based Plug-and-Play Method for Hyperspectral Image Reconstruction -- AGFNet: Attention Guided Fusion Network for Camouflaged Object Detection -- Spatial-Temporal Contextual Feature Fusion Network for Movie Description -- A Transformer-Based Network for Deformable Medical Image Registration -- Context Alignment Network for Video Moment Retrieval -- Underwater Object Detection Using Restructured SSD -- A Multi-Branch Cascade Transformer Network (MBCT-Net) for Hand Gesture Segmentation in Cluttered Background -- Relative Position Relationship Learning Network for Scene Graph Generation -- Person Re-Identification Based on CNN with Multi-Scale Contour Embedding -- Causal Inference With Sample Balancing for Out-Of-Distribution Detection in Visual Classification -- Low-Light Image Enhancement Under Mixed Noise Model With Tensor Representation -- Sequential Fusion of Multi-View Video Frames for 3D Scene Generation -- Multi-Model Lightweight Action Recognition with Group-Shuffle Graph Convolutional Network -- Global-View Re-Identification Tracking with Transformer -- VERTEX: VEhicle Reconstruction and TEXture Estimation From a Single ImageUsing Deep Implicit Semantic Template Mapping -- TAR-Net: A Triple Attention Residual Network for Power Line Extraction from Infrared Aerial Images -- Deep Dynamic-Range Compression of Infrared Video Camera -- Saliency Map-Based Local White-Box Adversarial Attack Against Deep Neural Networks -- Improving Adversarial Attacks With Ensemble-Based Approaches -- Browsing Behavioral Intent Prediction on Product Recommendation Pages of E-Commerce Platform -- An Optical Satellite Controller Based on Diffractive Deep Neural Network -- Incomplete Cigarette Code Recognition via Unified SPA Features and Graph Space Constraints -- A Large-Scale Tobacco 3D Bin Packing Model Based on Dual-Task Learning of Group Blocks -- ETH-TT: A Novel Approach For Detecting Ethereum Malicious Accounts -- Multi-Objective Meta-Return Reinforcement Learning for Sequential Recommendation -- Purchase Pattern Based Anti-Fraud Framework In Online E-Commerce Platform Using Graph Neural Network -- Physical Logic Enhanced Network for Small-Sample Bi-Layer Metallic Tubes Bending Springback Prediction -- Blind Surveillance Image Quality Assessment via Deep Neural Network Combined With the Visual Saliency -- Power Grid Bus Cluster Based on Voltage Phasor Trajectory -- Following the Lecturer: Hierarchical Knowledge Concepts Prediction for Educational Videos -- Learning Evidential Cognitive Diagnosis Networks Robust to Response Bias -- An Integrated Navigation Method for UAV Autonomous Landing Based on Inertial and Vision Sensors -- An Automatic Surface Defect Detection Method With Residual Attention Network -- Research on Intelligent Decision-Making Irrigation Model of Water and Fertilizer Based on Multi-Source Data Input -- Interaction-Aware Temporal Prescription Generation via Message Passing Neural Network -- Adversarial and Implicit Modality Imputation With Applications to Depression Early Detection -- EdgeVR360: Edge-Assisted Multiuser-Oriented Intelligent 360-Degree Video Delivery Scheme Over Wireless Network -- Automatic Diagnose of Drug-Resistance Tuberculosis From CT Images Based on Deep Neural Networks -- Maintaining Structural Information by Pairwise Similarity for Unsupervised Domain Adaptation -- Hierarchical Recurrent Contextual Attention Network for Video Question Answering -- An Intelligent Assessment System for Human Motor Functions of Stroke Patients -- UI Layers Group Detector: Grouping UI Layers via Text Fusion and Box Attention -- Audio-Visual Fusion Network Based on Conformer for Multimodal Emotion Recognition -- Audio-Visual Multi-Person Keyword Spotting via Hybrid Fusion -- A New Adaptive TV-Based BM3D Algorithm for Image Denoising -- Data-Driven Hybrid Neural Network under Model-Driven Supervised Learning for Structural Dynamic Impact Localization -- Connecting Patients With Pre-Diagnosis: A Multiple Graph Regularized Method for Mental Disorder Diagnosis -- A Novel Device-Free Localization Approach Based on Deep Dictionary Learning -- Survey of Hypergraph Neural Networks and Its Application to Action Recognition -- Visual Perception Inference on Raven's Progressive Matrices by Semi-Supervised Contrastive Learning -- PHN: Parallel Heterogeneous Network With Soft Gating for CTR Prediction -- Multi-Relational Cognitive Diagnosis for Intelligent Education -- Clinical Phenotyping Prediction via Auxiliary Task Selection and Adaptive Shared-Space Correction -- Multi-View Subspace Clustering With Joint Tensor Representation and Indicator Matrix Learning -- Weighted Competitive-Collaborative Representation Based Classifier for Imbalanced Data Classification -- Hierarchical Graph Representation Learning with Structural Attention for Graph Classification -- Photovolt Spots Detection Based on Kernel Entropy Component Analysis and Information Gain -- Second-Order Global Attention Networks for Graph Classification and Regression -- Aero-Engine Remaining Useful Life Prediction Via Tensor Decomposition Method -- FIGCI: Flow-Based Information-Geometric Causal Inference -- Emotional Semantic Neural RadianceFields for Audio-Driven Talking Head -- Multi-View 3D Morphable Face Reconstruction via Canonical Volume Fusion -- DPIT: Dual-Pipeline Integrated Transformer for Human Pose Estimation -- Integrative Analysis of Multi-view Histopathological Image Features for the Diagnosis of Lung Cancer -- Where Are the Children with Autism Looking in Reality? -- Study on Practical Utility of Image Dehazing

Algorithms Based on Deep Learning in Computer Vision Scene Understanding -- Benchmarking Deep Reinforcement Learning Based Energy Management Systems for Hybrid Electric Vehicles -- Self-Attention Based High Order Sequence Features of Dynamic Functional Connectivity Networks With rs-fMRI for Brain Disease Classification -- A New Method for Assigning Hesitation Based on An IFS Distance Metric -- Adaptive Combination of Filtered-x NLMS and Affine Projection Algorithms for Active Noise Control -- Linguistic Interval-Valued Spherical Fuzzy Sets and Related Properties -- A Genetic Algorithm for Causal Discovery Based on Structural Causal Model -- Stochastic and Dual Adversarial GAN-Boosted Zero-Shot Knowledge Graph -- LS-YOLO: Lightweight SAR Ship Targets Detection Based on Improved YOLOv5 -- Dictionary Learning-Based Reinforcement Learning With Non-Convex Sparsity Regularizer -- Deep Twin Support Vector Networks -- Region-Based Dense Adversarial Generation for Medical Image Segmentation -- Dynamic Clustering Federated Learning for Non-IID Data -- Dynamic Network Embedding by Using Sparse Deep Autoencoder -- Deep Graph Convolutional Networks Based on Contrastive Learning: Alleviating Over-Smoothing Phenomenon -- Clustering-Based Curriculum Construction for Sample-Balanced Federated Learning -- A Novel Nonlinear Dictionary Learning Algorithm Based on Nonlinear-KSVD and Nonlinear-MOD -- Tooth Defect Segmentation in 3D Mesh Scans using Deep Learning -- Crowd-Oriented Behavior Simulati.

This three-volume set LNCS 13604-13606 constitutes revised selected papers presented at the Second CAAI International Conference on Artificial Intelligence, held in Beijing, China, in August 2022. CICAI is a summit forum in the field of artificial intelligence and the 2022 forum was hosted by Chinese Association for Artificial Intelligence (CAAI). The 164 papers were thoroughly reviewed and selected from 521 submissions. CICAI aims to establish a global platform for international academic exchange, promote advanced research in AI and its affiliated disciplines such as machine learning, computer vision, natural language, processing, and data mining, amongst others.