Theory and Algorithms -- LSMVC: Low-rank Semi-supervised Multi-view Clustering for Special Equipment Safety Warning -- Single-Skeleton and Dual-Skeleton Hypergraph Convolution Neural Networks for Skeleton-Based Action Recognition -- Multi-Reservoir Echo State Network with Multiple-Size Input Time Slices for Nonlinear Time-Series Prediction -- Transformer with Prior Language Knowledge for Image Captioning -- Continual Learning with Laplace Operator based Node-Importance Dynamic Architecture Neural Network -- Improving generalization of reinforcement learning for multi-agent combating games -- Gradient Boosting Forest: A Two-Stage Ensemble Method Enabling Federated Learning of GBDTs -- Random Neural Graph Generation with Structure Evolution -- MatchMaker: Aspect-Based Sentiment Classification via Mutual Information -- PathSAGE: Spatial Graph Attention Neural Networks With Random Path Sampling -- Label Preserved Heterogeneous Network Embedding -- Spatio-Temporal Dynamic Multi-Graph Attention Network for Ride-hailing Demand Prediction -- An Implicit Learning Approach for Solving the Nurse Scheduling Problem -- Improving Goal-Oriented Visual Dialogue by Asking Fewer Questions -- Balance Between Performance and Robustness of Recurrent Neural Networks brought by Brain-inspired Constraints on Initial Structure -- Single-Image Smoker Detection by Human-Object Interaction with Post-Refinement -- A Lightweight Multi-scale Feature Fusion Network For Real-time Semantic Segmentation -- Multi-view Fractional Deep Canonical Correlation Analysis for Subspace Clustering -- Handling the Deviation from Isometry between Domains and Languages in Word Embeddings: Applications to Biomedical Text Translation -- Inference in Neural Networks Using Conditional Mean-Field Methods -- Associative Graphs for Fine-Grained Text Sentiment Analysis -- k-Winners-Take-All Ensemble Neural Network -- Performance Improvement of FORCE Learning for Chaotic Echo State Networks -- Generative Adversarial Domain Generalization via Cross-Task Feature Attention Learning for Prostate Segmentation -- Context-based Deep Learning Architecture with Optimal Integration Layer for Image Parsing -- Kernelized Transfer Feature Learning on Manifolds -- Data-Free Knowledge Distillation with Positive-Unlabeled Learning -- Manifold Discriminative Transfer Learning for Unsupervised Domain Adaptation -- Training-Free Multi-Objective Evolutionary Neural Architecture Search via Neural Tangent Kernel and Number of Linear Regions -- Neural Network Pruning via Genetic Wavelet Channel Search -- Binary Label-aware Transfer Learning for Cross-domain Slot Filling -- Condition-Invariant Physical Adversarial Attacks via Pixel-wise Adversarial Learning -- Multiple Partitions Alignment with Adaptive Similarity Learning -- Recommending best course of treatment based on similarities of prognostic markers -- Generative Adversarial Negative Imitation Learning from Noisy Demonstrations -- Detecting Helmets on Motorcyclists by Deep Neural Networks with aDual-Detection Scheme -- Short-Long Correlation Based Graph Neural Networks for Residential Load Forecasting -- Disentangled Feature Network for Fine-Grained Recognition -- Large-Scale Topological Radar Localization Using Learned Descriptors -- Rethinking binary hyperparameters for deep transfer learning -- Human Centred Computing -- Hierarchical Features Integration and Attention Iteration Network for Juvenile Refractive Power Prediction -- Stress Recognition in Thermal Videos using Bi-Directional Long-Term Recurrent Convolutional Neural Networks -- StressNet: A Deep Neural Network based on Dynamic Dropout Layers for Stress Recognition -- Analyzing Vietnamese Legal Questions using Deep Neural Networks with Biaffine Classifiers -- BenAV: A Bengali Audio-Visual Corpus for Visual Speech Recognition -- Investigation of Different G2P Schemes for Speech Recognition in Sanskrit -- GRU with Level-Aware Attention for Rumor Early Detection in Social Networks -- Convolutional Feature-interacted FactorizationMachines for Sparse Contextual Prediction -- A Lightweight Multidimensional Self-Attention Network for Fine-grained Action Recognition -- Unsupervised Domain Adaptation with Self-selected Active Learning for Cross-domain OCT Image Segmentation -- Adaptive Graph Convolutional Network with Prior Knowledge for Action Recognition -- Self-Adaptive Graph Neural Networks for Personalized Sequential Recommendation -- Spitial-Temporal Attention Network with Multi-Similarity Loss for Fine-Grained Skeleton-Based Action Recognition -- SRGAT: Social Relational Graph Attention Network for Human Trajectory Prediction -- FSE: A powerful feature augmentation technique for classification task -- AI and Cybersecurity -- FHTC: Few-shot Hierarchical Text Classification in Financial Domain -- JStrack: Enriching Malicious JavaScript Detection Based on AST Graph Analysis and Attention Mechanism.

The four-volume proceedings LNCS 13108, 13109, 13110, and 13111 constitutes the proceedings of the 28th International Conference on Neural Information Processing, ICONIP 2021, which was held during December 8-12, 2021. The conference was planned to take place in Bali, Indonesia but changed to an online format due to the COVID-19 pandemic. The total of 226 full papers presented in these proceedings was carefully reviewed and selected from 1093 submissions. The papers were organized in topical sections as follows: Part I: Theory and algorithms; Part II: Theory and algorithms; human centred computing; AI and cybersecurity; Part III: Cognitive neurosciences; reliable, robust, and secure machine learning algorithms; theory and applications of natural computing paradigms; advances in deep and shallow machine learning algorithms for biomedical data and imaging; applications; Part IV: Applications.