PRICAI 2022: Trends in Artificial Intelligence

<div>Recommender System.- Mixture of Graph Enhanced Expert Networks for Multi-task&nbsp;Recommendation.-&nbsp;MF-TagRec: Multi-feature fused tag recommendation for GitHub.-&nbsp;Co-contrastive Learning for Multi-behavior Recommendation.-&nbsp;Pattern Matching and Information-aware between Reviews and&nbsp;Ratings for Recommendation.-&nbsp;Cross-view&nbsp;Contrastive Learning for Knowledge-aware Session-based Recommendation.-&nbsp;Reinforcement Learning.- HiSA: Facilitating Efficient Multi-Agent Coordination and Cooperation&nbsp;by Hierarchical Policy with Shared Attention.- DDMA:&nbsp;Discrepancy-Driven Multi-Agent Reinforcement Learning.-&nbsp;PRAG: Periodic Regularized Action Gradient for Efficient Continuous&nbsp;Control.-&nbsp;Identifying Multiple Influential Nodes for&nbsp;Complex Networks based on Multi-Agent Deep Reinforcement Learning.-&nbsp;Online Learning in Iterated Prisoner’s Dilemma to Mimic&nbsp;Human Behavior.- Optimizing&nbsp;Exploration-Exploitation Trade-off in Continuous Action Spaces via&nbsp;Q-ensemble.-&nbsp;Hidden&nbsp;Information General Game Playing With Deep Learning and Search.-&nbsp;Sequential Decision&nbsp;Making with “Sequential Information” in Deep Reinforcement Learning.-&nbsp;Two-Stream&nbsp;Communication-Efficient Federated Pruning Network.-&nbsp;Strong General AI.- Multi-scale Lightweight Neural Network for Real-time Object Detection.- Hyperspectral&nbsp;Image Classification Based On Transformer and Generative Adversarial&nbsp;Network.-&nbsp;Deliberation Selector for Knowledge-grounded Conversation Generation.-&nbsp;Training a Lightweight ViT Network for Image Retrieval.-&nbsp;Vision and Perception.- Segmented–original Image Pairs to Facilitate Feature Extraction in&nbsp;Deep Learning Models.- FusionSeg: Motion Segmentation&nbsp;by Jointly Exploiting Frames and Events.-&nbsp;Weakly-supervised Temporal Action Localization&nbsp;with Multi-head Cross-modal Attention.-&nbsp;CrGAN:&nbsp;Continuous Rendering of Image Style.-&nbsp;DPCN: Dual Path Convolutional&nbsp;Network for Single Image Deraining.-&nbsp;All Up&nbsp;to You: Controllable Video Captioning With a Masked Scene Graph.-&nbsp;A Multi-Head Convolutional Neural&nbsp;Network With Multi-path Attention improves Image Denoising.-&nbsp;Learning&nbsp;Spatial Fusion and Matching for Visual Object Tracking.-&nbsp;Lightweight Wavelet-based Transformer&nbsp;for Image Super-resolution.- Efficient high-resolution human pose&nbsp;estimation.-&nbsp;The&nbsp;Geometry Enhanced Deep Implicit Function based 3D Reconstruction&nbsp;for objects in a real-scene image.-&nbsp;Multi-View Stereo Network with Attention Thin Volume.- 3D Point Cloud Segmentation Leveraging Global 2D-view Features.-&nbsp;Self-Supervised Indoor&nbsp;360-Degree Depth Estimation via Structural Regularization.-&nbsp;Global Boundary Refinement for Semantic&nbsp;Segmentation via Optimal Transport.-&nbsp;Optimization-based Predictive Approach for On-Demand&nbsp; Transportation.- Joint Contrast: Skeleton-based Mutual Action&nbsp;Recognition with Contrastive Learning.-&nbsp;Nested Multi-Axis Learning Network for Single Image&nbsp;Super Resolution.-&nbsp;Efficient Scale Divide and&nbsp;Conquer Network for Object Detection.-&nbsp;Video-Based Emotion Recognition in the Wild for&nbsp;Online Education Systems.-&nbsp;Real-world Underwater Image Enhancement via Degradation-aware&nbsp;Dynamic Network.-&nbsp;Self-Supervised Vision Transformer based Nearest Neighbor&nbsp;Classification for Multi-Source Open-Set Domain Adaptation.-&nbsp;Lightweight image dehazing&nbsp;neural network model based on estimating medium transmission map by intensity.- CMNet: Cross-aggregation Multi-branch Network for&nbsp;Salient Object Detection.-&nbsp;More than Accuracy: an Empirical Study of Consistency&nbsp;between Performance and Interpretability.-&nbsp;Object-scale Adaptive Optical Flow Estimation Network.- A Task-aware Dual&nbsp;Similarity Network for Fine-grained Few-shot Learning.-&nbsp;Rotating Target&nbsp;Detection Based On Lightweight Network.-&nbsp;Corner Detection Based on a Dynamic Measure of Cornerity.</div>