Automatic speech recognition (ASR) systems have achieved remarkable performance in common conditions but often struggle to leverage long-context information in contextualized scenarios that require domain-specific knowledge, such as conference presentations. This challenge arises primarily due to constrained model context windows and the sparsity of relevant information within extensive contextual noise. To solve this, we propose the SAP^2 method, a novel framework that dynamically prunes and integrates relevant contextual keywords in two stages. Specifically, each stage leverages our proposed Speech-Driven Attention-based Pooling mechanism, enabling efficient compression of context embeddings while preserving speech-salient information. Experimental results demonstrate state-of-the-art performance of SAP^2 on the SlideSpeech and LibriSpeech datasets, achieving word error rates (WER) of 7.71% and 1.12%, respectively. On SlideSpeech, our method notably reduces biased keyword error rates (B-WER) by 41.1% compared to non-contextual baselines. SAP^2 also exhibits robust scalability, consistently maintaining performance under extensive contextual input conditions on both datasets.

Stress is an inevitable aspect of modern life. Real-time stress detection methods can assist individuals in regulating their mental states and enhancing psychological well-being. This study utilized data from over 180 participants, who were divided into biological stressor, social stressor, and control groups. Physiological indicators included respiratory sinus arrhythmia (RSA), skin conductance level (SCL), and cortisol levels. Zdevelop new algorithms on low-frequency wearable physiological data. Considering the problem of high noise in this type of data, we first proposed and verified the effectiveness of the algorithm on high-frequency physiological instrument data, and then transplanted the effective algorithm to low-frequency wearable data. Ultimately, this paper provides a comprehensive dataset covering various types of stressors and demonstrates encouraging results in stress detection, achieving a precision of 83.3% on watch data.

Automatic Speech Recognition (ASR) models pre-trained on large-scale speech datasets have achieved significant breakthroughs compared with traditional methods. However, mainstream pre-trained ASR models encounter challenges in distinguishing homophones, which have close or identical pronunciations. Previous studies have introduced visual auxiliary cues to address this challenge, yet the sophisticated use of lip movements falls short in correcting homophone errors. On the other hand, the fusion and utilization of scene images remain in an exploratory stage, with performance still inferior to the pre-trained speech model. In this paper, we introduce CIEASR (Contextual Image-Enhanced Automatic Speech Recognition), a novel multimodal speech recognition model that incorporates a new cue fusion method, using scene images as soft prompts to correct homophone errors. To mitigate data scarcity, we refine and expand the VSDial dataset for extensive experiments, illustrating that scene images contribute to the accurate recognition of entity nouns and personal pronouns. Our proposed CIEASR achieves state-of-the-art results on VSDial and Flickr8K, significantly reducing the Character Error Rate (CER) on VSDial from 3.61% to 0.92%.

Ultrasound image guidance could display the movement of soft tissue in real time, which provides an important basis for the operation path selection of tumor for examinations and interventions such as precise localization and puncture biopsy. However, factors such as tissue deformation, patient motion, and instrument contact pose great challenges for real-time smooth localization tracking of target tissues and maintaining a suitable acoustic window. Not only does it reduce the accuracy of ultrasound examination, but also may prolong the examination time. In this paper, a real-time autonomous ultrasound robot trajectory planning framework is proposed to achieve global scanning of breast tissue and real-time local tracking of suspicious tumor targets. In addition, an ultrasound suspicious tumor target next moment motion attitude and position estimation algorithm and an acceleration-continuous online trajectory generation (ACOTG) algorithm are proposed. Experiments on breast phantom showed that the fluctuation difference of both global scanning and tracking scanning force was less than 5 N$\pm$15%, and ACOTG takes less than 1 microsecond to calculate the next moment state. The position, velocity, and acceleration of the tracking path did not change abruptly when the unknown target changed. The ultrasound image of the suspected tumor target could always be smoothly maintained in a suitable acoustic window, while the errors between the center of the suspected tumor target image and the center of the acoustic window in the horizontal direction and the depth direction were less than$\pm$1 mm and 30 mm$\pm$10%, respectively.Note to Practitioners—The motivation of this study was to solve the clinical problem of difficulty in keeping the suspected breast tumor target stably in the center of the field of view under the influence of tissue deformation, patient motion and instrument contact in ultrasound guidance. We proposed the idea of global conventional scanning and local online tracking to achieve the autonomous scanning, detecting and tracking of suspicious tumor targets within the global scope of the breast. ACOTG algorithm is proposed and implemented, which can make the ultrasound image of suspicious tumor target can be displayed stably in the center of the acoustic window. With the help of this system, the accuracy of ultrasound examination can be improved and the time of examination can be reduced. The proposed system can also be applied to ultrasound scanning of other parts of the body.

The quality of breast ultrasound imaging is greatly affected by the contact force of the probe, which largely requires experienced sonographers to complete the clinical examination. We propose a flexible and fully autonomous ultrasound scanning system for breast ultrasound imaging. It consists of an ultrasound machine, a dual robotic arms system, a multi-structured light system, a human–computer interaction system, and a flexible ultrasound probe clamping device (FUPCD). First, the dynamics model of the FUPCD was analyzed, and a closed-loop force control strategy was established. We then implemented an automatic scanning system. The hysteresis characteristics of the FUPCD and transient response of the force controller were experimentally verified. The system could keep the steady-state error less than ± 5% within 0.5 s. Second, the performance of the control system to maintain constant contact force at different scanning speeds (Note to Practitioners—The motivation of this study is to solve the problem of poor image reproducibility in breast ultrasound scanning, but the proposed system can also be applied to ultrasound scanning of other parts of the body. The position, direction, and contact force of the ultrasound probe affects the image quality and repeatability of the ultrasound, thereby affecting the diagnostic ability. Therefore, we propose and develop a flexible and fully autonomous breast ultrasound scanning system. We aimed to improve the autonomy and stability of the scanning process by designing end-to-end automated scanning strategies, flexible clamping devices, and closed-loop force control strategies. Among them, the fully automatic three-dimensional perception and trajectory planning can provide global fitting capabilities to different forms of breast surfaces and control the probe to maintain the best contact posture. At the same time, the fully automatic workflow reduces additional interference and reduces the complexity of the workflow. The flexible ultrasound probe clamping device improves the passive applicability to flexible tissues and reduces the resistance in the scanning process. The closed-loop force control strategy can adjust the contact force in real time. Experiments verified that the repeatability could be evaluated by contact force. A series of human–machine comparison experiments verified the advancement and effectiveness of the system. In future research, we will further solve the problem of abnormalities and repeatability of ultrasound images acquired during the scanning process by combining the multi-modal feedback of images and forces.

Logistics interfaces with manufacturing throughout the entire production process need synchronous operations. For achieving integrated organization and operations between manufacturing and logistics, this article introduces the concept of shop-floor logistics and manufacturing synchronization with four principles, including: 1) synchronization-oriented manufacturing system; 2) synchronized information sharing; 3) synchronized decision making; and 4) synchronized operations. The marriage of the principles rendered the development of an overall framework of the Industrial Internet of Things (IIoT) and digital twin-enabled graduation intelligent manufacturing system (GiMS). A mixed-integer programming-based synchronization mechanism is proposed under GiMS. To meet the requirement of fast decision making in real-life shop-floor logistics and manufacturing synchronization problems, an equivalent constraint programming model is developed and tested. The observation and analysis of the case company show the advantage of the proposed concept and approach with the best performance regarding key performance indicators. The concept of synchronization provides an insight for understanding the interaction of logistics and manufacturing at the operational level. This article potentially enables manufacturers to reevaluate and develop their manufacturing planning and control strategies in the IIoT and digital twin-based manufacturing environment.

Visual information is indispensable to human locomotion in complex environments. Although amputees can perceive the environmental information by eyes, they cannot transmit the neural signals to prostheses directly. To augment human-prosthesis interaction, this article introduces a subvision system that can perceive environments actively, assist to control the powered prosthesis predictively, and accordingly reconstruct a complete vision-locomotion loop for transfemoral amputees. By using deep learning, the subvision system can classify common static terrains (e.g., level ground, stairs, and ramps) and estimate corresponding motion intents of amputees with high accuracy (98%). After applying the subvision system to the locomotion control system, the powered prosthesis can help amputees to achieve nonrhythmic locomotion naturally, including switching between different locomotion modes and crossing the obstacle. The subvision system can also recognize dynamic objects, such as an unexpected obstacle approaching the amputee, and assist in generating an agile obstacle-avoidance reflex movement. The experimental results demonstrate that the subvision system can cooperate with the powered prosthesis to reconstruct a complete vision-locomotion loop, which enhances the environmental adaptability of the amputees.

Interoperable manufacturing systems help manufacturing companies stay competitive in the environment of frequent and unpredictable market changes. An important part of a manufacturing system is computer numerically controlled (CNC) machine tools. Over the years, G-codes have been extensively used by CNC machine tools and are now considered as a bottleneck for making these machines adaptable and interoperable. Two new technologies emerged in recent years: Standard for the Exchange of Product data for Numerical Control (STEP-NC) and function blocks. The STEP-NC data model represents a common standard for NC programming, making the goal of a generic NC code generation facility a reality. Function blocks are an emerging IEC standard for distributed industrial processes and control systems. They can be used for CNC controls to encapsulate machining data, such as machining features and their needed algorithms. This paper introduces the above two new standards and the technologies that are developed based on the standards. The main body is devoted to analyze the standards from the functionality viewpoint. These functionalities include, bidirectional information flow in computer-aided design/computer-aided manufacturing, data sharing over the Internet, the use of feature-based machining concept, modularity and reusability, intelligent and autonomous CNC, and portability among resources. Some implementations are also presented to showcase how the standards are used to develop technologies for interoperable machining. Note to Practitioners-Modern computer numerically controlled (CNC) machine tools are limited in functions because their controllers rely on G-codes for communications. G-code is considered a "dumb" language as it only documents instructional and procedural data, leaving most of the design information behind. G-code programs are also hardware dependent, denying modern CNC machine tools desired interoperability and portability. In recent years, two new standards emerged, STEP-NC and function blocks. They may hold the key to empowering CNC machine tools with richer information which, in turn, gives CNC machine tools the ability to "think" intelligently and to be interoperable. This paper introduces these two standards, the technologies that have been developed based on the standards and some prototype systems using the standards and technologies. The intention is not to highlight any achieved research outcome. Instead, the focus is on informing the research and practical world about these new standards, analyzing them from the viewpoint of supporting interoperable CNC machine tools, and offering some futuristic views about these standards and technologies. While these standards are still in their infancy, research activities and prototype systems are already coming thick and fast. There seems to be a "healthy" mixture of participants working in the field. They range from the manufacturers of all systems related to the data interface (i.e., CAM systems, controls, and machine tools), to the users and academic institutions.