Recording the trajectory of table tennis balls in real-time enables the analysis of the opponent’s attacking characteristics and weaknesses. The current analysis of the ball paths mainly relied on human viewing, which lacked certain theoretical data support. In order to solve the problem of the lack of objective data analysis in the research of table tennis competition, a target detection algorithm-based table tennis trajectory extraction network was proposed to record the trajectory of the table tennis movement in video. The network improved the feature reuse rate in order to achieve a lightweight network and enhance the detection accuracy. The core of the network was the “feature store & return” module, which could store the output of the current network layer and pass the features to the input of the network layer at the next moment to achieve efficient reuse of the features. In this module, the Transformer model was used to secondarily process the features, build the global association information, and enhance the feature richness of the feature map. According to the designed experiments, the detection accuracy of the network was 96.8% for table tennis and 89.1% for target localization. Moreover, the parameter size of the model was only 7.68 MB, and the detection frame rate could reach 634.19 FPS using the hardware for the tests. In summary, the network designed in this paper has the characteristics of both lightweight and high precision in table tennis detection, and the performance of the proposed model significantly outperforms that of the existing models.

Recognizing behaviors within classroom settings is vital for gauging educational progress and optimizing teaching methodologies. The complexity of classroom environments often poses challenges for detectors in terms of detection performance. Despite having a simplified structure for analyzing actions in classroom environments, detection transformer (DETR) encounters certain challenges. To overcome these obstacles, we propose a novel real-time end-to-end object detection model called PACR-DETR. Specifically, we introduce an efficient position-aware channel residual network (PACRNet), which excels at extracting information from deep layers along the spatial dimension within complex backgrounds. Notably, it exclusively employs the transformer encoder to process the deepest layer of the feature map. Afterwards, we introduce the efficient residual mixing block (ERM Block), in order to apply convolution operations selectively to a subset of channels. This block underpins the multi-branch perception cross-level fusion module (MPCF Module), which facilitates hierarchical and cross-layer information fusion. Through a bidirectional chain, the multi-branch perception cross-level fusion network (MPCFNet) is architected to comprehensively integrate multi-scale gradient information. Finally, we integrate the normalized Wasserstein distance (NWD) with the Inner-MPDIoU loss function to enhance the recognition performance for small-scale objects. The results show that the PACR-DETR outperforms other methods in detection performance and real-time capability. On the SCB03-S dataset for classroom behavior detection against complex backgrounds, the proposed method attains a precision of 72.1%, outperforming YOLOv6 by 14.1% and YOLOv8m by 4.5%. Furthermore, it reaches 41.67 FPS, meeting the requirements for real-time application.

Under a student-centered educational paradigm, project-based learning (PBL) assessment requires accurate identification of classroom behaviors to facilitate effective teaching evaluations and the implementation of personalized learning strategies. The increasing use of visual and multi-modal sensors in smart classrooms has made it possible to continuously capture rich behavioral data. However, challenges such as lighting variations, occlusions, and diverse behaviors complicate sensor-based behavior analysis. To address these issues, we introduce CSSA-YOLO, a novel detection network that incorporates cross-scale feature optimization. First, we establish a C2fs module that captures spatiotemporal dependencies in small-scale actions such as hand-raising through hierarchical window attention. Second, a Shuffle Attention mechanism is then integrated into the neck to suppress interference from complex backgrounds, thereby enhancing the model’s ability to focus on relevant features. Finally, to further enhance the network’s ability to detect small targets and complex boundary behaviors, we utilize the WIoU loss function, which dynamically weights gradients to optimize the localization accuracy of occluded targets. Experiments involving the SCB03-S dataset showed that CSSA-YOLO outperforms traditional methods, achieving an mAP₅₀ of 76.0%, surpassing YOLOv8m by 1.2%, particularly in complex background and occlusion scenarios. Furthermore, it reaches 78.31 FPS, meeting the requirements for real-time application. This study offers a reliable solution for precise behavior recognition in classroom settings, supporting the development of intelligent education systems.

Advances in deep learning and computer vision have overcome many challenges inherent in the field of autonomous intelligent vehicles. To improve the detection accuracy and efficiency of EdgeBoard intelligent vehicles, we proposed an optimized design of EdgeBoard based on our PP-YOLOE+ model. This model innovatively introduces a composite backbone network, incorporating deep residual networks, feature pyramid networks, and RepResBlock structures to enrich environmental perception capabilities through the advanced analysis of sensor data. The incorporation of an efficient task-aligned head (ET-head) in the PP-YOLOE+ framework marks a pivotal innovation for precise interpretation of sensor information, addressing the interplay between classification and localization tasks with high effectiveness. Subsequent refinement of target regions by detection head units significantly sharpens the system’s ability to navigate and adapt to diverse driving scenarios. Our innovative hardware design, featuring a custom-designed mainboard and drive board, is specifically tailored to enhance the computational speed and data processing capabilities of intelligent vehicles. Furthermore, the optimization of our Pos-PID control algorithm allows the system to dynamically adjust to complex driving scenarios, significantly enhancing vehicle safety and reliability. Besides, our methodology leverages the latest technologies in edge computing and dynamic label assignment, enhancing intelligent vehicles’ operations through seamless sensor integration. Our custom dataset, specifically designed for this study, includes 4777 images captured by intelligent vehicles under a variety of environmental and lighting conditions. The dataset features diverse scenarios and objects pertinent to autonomous driving, such as pedestrian crossings and traffic signs, ensuring a comprehensive evaluation of the model’s performance. We conducted extensive testing of our model on this dataset to thoroughly assess sensor performance. Evaluated against metrics including accuracy, error rate, precision, recall, mean average precision (mAP), and F1-score, our findings reveal that the model achieves a remarkable accuracy rate of 99.113%, an mAP of 54.9%, and a real-time detection frame rate of 192 FPS, all within a compact parameter footprint of just 81 MB. These results demonstrate the superior capability of our PP-YOLOE+ model to integrate sensor data, achieving an optimal balance between detection accuracy and computational speed compared with existing algorithms.

In view of the high demand for low-power biomedical signal-processing schemes occupying very less on-chip areas, we propose a low-power highly integrated system-on-chip (SoC) design comprising a nonuniform biomedical sensing signal conditioning circuit and a 32-bit microprocessor. The proposed design is implemented using a 0.18 µm CMOS EFLASH process in an area of 4.2 mm², and is applied in PPG, ECG and BP measurements. Experimental results show that our SoC design provides extremely good performances for sparse features of all three biomedical applications. The charge consumption is 25 mAh for three months, for an experimental board design including sleep current and junction leakage current.

In this paper, a 6-bit high-speed digital-to-analog converter (DAC) is presented. This DAC is based on a segmented architecture and has an operating speed up to 4GS/s according to the post-layout simulation results. The output waveform of this DAC is realized in a non-return-to-zero (NRZ) way. The DAC core occupies an area of 0.09mm² in a 40nm CMOS technology. A DfT block is introduced to relieve the speed requirement of high-speed I/O. The spurious free dynamic range (SFDR) up to 44.81dBc is achieved over Nyquist interval. The power consumption is 13mW at near Nyquist frequency.