Objective: To construct a knowledge graph of acupuncture case studies from renowned acupuncturists, analyze acupuncture medical case knowledge, mine implicit knowledge, and perform visual representation, providing methodological references for the research of acupuncture medical cases. Methods: The types of knowledge entities and relationships between entities involved in acupuncture case studies from renowned acupuncturists were sorted out. The TCM Miner platform was used for text annotation, and a large language model was employed to identify and label entities. The relationships between entities were determined based on Chinese medicine-related standards, literature reviews, and established rules. After knowledge fusion, the data was imported into the Neo4j graph database using the Cypher language for storage and visual representation, constructing a knowledge graph of acupuncture medical cases. Results: The acupuncture medical case knowledge graph contained 577 nodes and 3,905 relationships, with a schema comprising 5 entity classes and 5 relationship types. Through Cypher language queries, knowledge can be visually presented in four aspects: commonly used treatment methods of practitioners, research on the relationship between acupuncture techniques and diseases, exploration of disease-acupoint patterns, and research on other therapies and diseases. Discussion: The knowledge graph constructed in this study can visually display the knowledge and implicit relationships recorded in acupuncture medical cases, making it suitable for knowledge mining and visual representation of acupuncture medical cases.

This article studies the event-triggered adaptive fuzzy output feedback fault-tolerant control problem for nonaffine uncertain systems with output tracking errors constraints and actuator faults. First, the mean value theorem and input compensation method are used to decouple the control input signal, and an adaptive state observer is established to estimate unknown state. Then, in the presence of disturbances and actuator faults, an improved event-triggered adaptive fault-tolerant control strategy is designed to adjust control input signals, which effectively reduce the computing burden in communication, and the input-to-state stability assumption is not required. Furthermore, it is proved by Lyapunov stability analyses that the proposed control method guarantees output tracking errors can converge to the prescribed performance bounds and all signals are uniformly ultimately bounded. Finally, the validity of the control algorithm is verified through the simulation results.

This article considers the problem of finite-time adaptive fuzzy switching event-triggered control for a class of nonaffine stochastic systems with periodic actuator faults and asymmetric error constraints. First, a framework of semiglobally finite-time stability in probability is established for stochastic systems. Then, unlike the existing observers, an adaptive state observer with faults compensation mechanism is designed to estimate unknown state. In order to balance the systems tracking performance and communication burdens, a switching event-triggered strategy is given to regulate trigger signal and avoids the Zeno behavior effectively. In backstepping process, a nonlinear tracking error-dependent function is constructed to constrain the tracking errors within asymmetric boundaries, and the effects of faults and trigger errors can be compensated completely by the designed finite-time fault-tolerant control strategy. Finally, the effectiveness of presented strategy is further verified by simulation results.

Multi-human parsing aims to segment every body part of every human instance. Nearly all state-of-the-art methods follow the "detection first" or "segmentation first" pipelines. Different from them, we present an end-to-end and box-free pipeline from a new and more human-intuitive perspective. In training time, we directly do instance segmentation on humans and parts. More specifically, we introduce a notion of "indiscriminate objects with categorie" which treats humans and parts without distinction and regards them both as instances with categories. In the mask prediction, each binary mask is obtained by a combination of prototypes shared among all human and part categories. In inference time, we design a brand-new grouping post-processing method that relates each part instance with one single human instance and groups them together to obtain the final human-level parsing result. We name our method as Nondiscriminatory Treatment between Humans and Parts for Human Parsing (NTHP). Experiments show that our network performs superiorly against state-of-the-art methods by a large margin on the MHP v2.0 and PASCAL-Person-Part datasets.

This study proposes a new method for content‐based image retrieval by finding an optimal classifier. The optimal classifier is achieved by a new active learning support vector machine (SVM) which combines the model selection with the active learning. The unlabelled samples close to the boundary of the SVM classifier are selected based on the feature similarity for the active learning, and the adaptive regularisation is used to select the optimal model. The combination of model selection with active learning accelerates the convergence of the classifier. The new method can improve the image retrieval accuracy and reduce the time consumption. The experimental results show that the proposed method has a better performance with fewer samples and less time consumption for image retrieval.