Object detection and recognition in drone aerial images hold broad application value, but also present challenges such as large variations in object scales, difficulties in detecting small objects, and occlusions in dense scenes. To address these issues, this paper proposes an improved object detection algorithm based on RT-DETR. First, a Spatial-Channel Collaborative Attention (SCSA) module is introduced into the PResNet backbone network to enhance feature representation and improve detection accuracy. Second, the Content-Aware ReAssembly of Features (CARAFE) upsampling method is adopted in the Hybrid Encoder, which preserves more detailed information of small objects while reducing model complexity, further boosting detection performance. Finally, a modified MFRC3 module incorporating Biphasic Feature Aggregation Module (BFAM) and boundary attention mechanism is proposed to replace the original CSPRepLayer. This enhances multi-scale feature fusion and improves the retention of fine-grained and textural features.Experimental results on the VisDrone2019 datasets show that the improved algorithm achieves an mAP@0.5 of 51.1%, which is 3.1% higher than the baseline RT-DETR model.

Driven by deep learning advances, speech processing systems such as automatic speech recognition (ASR), source segregation, noise suppression have achieved significant performance improvements. However, traditional training strategies, particularly static mini-batch selection, often overlook the dynamic variations in data complexity and model convergence behavior, resulting in ineffective training efficiency and limited model accuracy. To tackle this limitation, we introduce a novel training paradigm called Dynamic Micro-block Optimization (DMBO). The method introduces a fine-grained sampling mechanism by partitioning the training set into smaller units called “micro-blocks,” which are dynamically updated during training based on real-time characteristics such as sample loss, gradient diversity, and utterance complexity. Four sampling strategies—loss-weighted, gradient-diversity, gender-based, and accent-based—are designed to self-adjust the composition of training data. The DMBO framework is implemented using Connectionist Temporal Classification (CTC) and Long Short-term Memory (LSTM) networks for end-to-end speech recognition. Experimental evaluations on the VCTK datasets demonstrate that the proposed method significantly accelerates convergence and improves model accuracy. Specifically, the gender-homogeneous strategy reduces the Label Error Rate (LER) by 9.0% compared to standard mini-batch training, while the accent-heterogeneous strategy achieves a 9.2% absolute LER reduction. These results confirm that dynamic optimization at the micro-block level enhances the efficacy of deep learning models in speech processing tasks, and the experimental outcomes are consistent with theoretical expectations, validating the effectiveness and correctness of the proposed approach.

Polar codes, as capacity-achieving error-correcting codes, have become a cornerstone of modern communication systems due to their excellent theoretical performance. Compared with the Successive Cancellation (SC) decoding algorithm, the Belief Propagation (BP) decoding algorithm for polar codes offers advantages such as parallel output and ease of hardware implementation. However, the bit-flipping decoding schemes based on BP still exhibit a significant performance gap in frame error rate (FER) compared to the Successive Cancellation List (SCL) decoding. To address the demand for high reliability and low power consumption in practical applications, this paper proposes an optimized bit-flipping scheme in which the flipping set is constructed using an adaptive genetic algorithm. The proposed method first reduces the computational complexity of the initial BP decoding process by adopting the Offset Min-Sum (OMS) approximation. During the construction of the flipping set, an adaptive mechanism dynamically adjusts the crossover and variational probabilities based on the fitness of individuals in the population. The indices of the information bits are used as individuals in the genetic algorithm, enabling the fitness values to gradually evolve from local optima toward a global optimum. This approach allows for more accurate identification of bit positions prone to decoding errors. For a polar code with a length of 1024 and a code rate of 0.5, the proposed AGA-OMS-BPF decoder achieves approximately 1.3 dB BER performance gain at a BER of 10⁻⁵ compared with the conventional BPF decoder. Simulation results demonstrate that the proposed method effectively reduces the number of unsuccessful BP decoding attempts by constructing a more efficient flipping set, thereby achieving performance gains with reduced computational complexity.

Addressing the common challenges in night vision imagery—poor lighting conditions, low pixel resolution, and diminished contrast—which hinder effective pedestrian feature extraction and result in suboptimal accuracy and real-time performance for night-time pedestrian detection, This paper proposes a deep learning-based night vision pedestrian detection system. Building upon the YOLOv8 object detection algorithm, the model is enhanced by incorporating the CBAM attention mechanism into its network architecture and upgrading the optimiser from SGD to Lion. The system design and development are further tailored to address the specific characteristics of night-time imagery. After experimental simulation verification, the performance of the improved algorithm model has been significantly improved: the overall accuracy is improved by about 2.0%, mAP@0.5 is improved by 1.6%, the average accuracy of IoU threshold 0.5 to 0.95 is improved by about 0.04%, and the F1 Score is improved by 0.64%. The improvement plan proposed in this paper effectively enhances the model's comprehensive identification ability of night vision pedestrians, improves the overall performance of the system, and verifies the correctness and validity of the research.

Human beings rely primarily on vision to perceive and interact with the external world, with approximately 80% of sensory information input through the visual system. This visual dominance makes the question of "where an individual is looking" not only a key to understanding attention distribution and information processing mechanisms but also a critical factor in optimizing decision-making efficiency and learning outcomes. However, traditional methods for analyzing gaze-related behaviors—such as manual behavioral observation and self-reported evaluation—suffer from inherent limitations: be havioral observation relies on subjective judgment of observers, often missing subtle gaze shifts and failing to achieve real-time tracking; self-evaluation is prone to memory biases and social desirability effects, leading to deviations between reported and actual gaze patterns. These drawbacks highlight the need for a more objective and precise alternative.Gaze estimation, which infers an individual’s visual attention and behavioral intentions by recording and analyzing the spatial position, movement trajectory, and dynamic changes of the eyeball, emerges as an ideal solution. This technology is broadly categorized into model-based (relying on geometric eye models) and appearance-based (using facial/ocular image features) approaches, with appearance-based methods gaining traction due to their non-intrusiveness. Nevertheless, current appearance-based gaze estimation still faces two major challenges: (1) individual differences, such as variations in eye shape, pupil size, eyelid structure, and the presence of glasses, which disrupt consistent feature extraction; (2) environmental interference, including variable lighting, partial facial occlusion, and dynamic head poses, which reduce estimation accuracy. To address these issues, this paper proposes RTACM-Net, a novel gaze estimation network architecture that integrates the strengths of Vision Transformer (ViT) with a multi-scale feature fusion mechanism. Specifically, RTACM-Net employs a lightweight convolutional module to extract local fine-grained features of the ocular region, while leveraging ViT’s multi-head attention mechanism to capture global contextual relationships. This dual-branch design enables the network to balance local feature precision and global context awareness, thereby mitigating the impact of individual differences and environmental noise.Extensive experiments were conducted on two benchmark datasets: MPIIFaceGaze (a large-scale dataset focusing on indoor controlled environments with 21 subjects) and Gaze360 (a challenging dataset covering diverse outdoor/indoor scenes, variable lighting, and large head-pose variations with over 100 subjects). The results show that RTACM-Net : on MPIIFaceGaze, it achieves an average angular error (MAE) of 3.72°; on Gaze360, it achieves an MAE of 10.46°, Gaze360-Net (11.40°) by 0.94°. These results demonstrate the robustness of RTACM-Net in handling variable individual characteristics and complex environmental conditions. Its practical potential extends to multiple fields: in augmented reality (AR), it can enable adaptive interface rendering; in autonomous driving, it supports dual-task monitorin; in human-robot interaction, it facilitates intuitive service triggering.

The rapid advancement of mobile technologies has led to increasingly powerful and feature-rich devices, yet this progress has also intensified the challenge of managing energy consumption effectively. Power optimization has therefore become a critical focus in mobile operating systems (OS), aiming to balance performance, functionality, and energy efficiency. As mobile devices integrate more complex hardware components and resource-intensive applications, ensuring sustainable power usage has become essential for improving battery life, user experience, and environmental sustainability. This research explores the fundamental question: How can mobile operating systems intelligently manage hardware and software resources to minimize power consumption without compromising performance or usability? To address this, the study examines key power optimization strategies and mechanisms integrated within modern mobile OS architectures, including Dynamic Voltage and Frequency Scaling (DVFS), power-aware CPU scheduling, Doze and App Standby modes, adaptive display and sensor management, and network optimization. The research also investigates the role of advanced techniques such as context-aware power management and machine learning-based predictive models in achieving dynamic, intelligent energy control. Using tools like Trepn Profiler, PowerTutor, and Android Battery Historian, the study evaluates how power consumption patterns can be analyzed and optimized in real time. The findings reveal that combining hardware-level techniques (like voltage scaling and clock gating) with software-level optimizations (such as adaptive scheduling and contextual awareness) results in significant energy savings while maintaining user satisfaction. Furthermore, the study highlights emerging challenges, including the trade-offs between performance and energy efficiency, the integration of AI for predictive optimization, and the need for sustainability across the device lifecycle. Ultimately, this research demonstrates that power optimization in mobile operating systems is not merely a technical requirement but a cornerstone of sustainable computing. Through intelligent power management, future mobile OSs can achieve greater efficiency, extended device longevity, and reduced environmental impact aligning technological innovation with eco-efficiency and user-centric design.