Irregular applications, such as graph analytics and sparse linear algebra, exhibit frequent indirect, data-dependent accesses to single or short sequences of elements that cause high main memory traffic limit performance. Data compression is a promising way accelerate irregular applications by reducing traffic. However, software adds substantial overheads, prior hardware techniques work poorly on the complex access patterns applications.We present SpZip, an architectural approach makes data...

Existing automatic parking path planning methods usually use circular curves and lines. At the junction of paths, are not smooth. Based on a low-speed kinematic model vehicle, this study redesigned paths. Sixth-degree polynomial based scattering points easement curve used in planning. two curved segments, paths continuous original, first-order second-order equations. Script programming was MATLAB to draw area swept by vehicle. Parking simulations were conducted using Carsim demonstrate...

Traditional computer vision methods are used in the urban road environment, but they vulnerable to environmental information interference and have low detection accuracy. To solve this problem, a multi-sensor method with strong anti-interference ability desired accuracy is proposed paper. In method, TensorFlow SSD-based deep learning determine approximate position of parking space on panoramic image marked rectangular box. box, grayscale, filtering, morphological open operation, binarization...

The key to traction control in locomotive systems is adjusting the axle torque through adhesion control. In traditional systems, due inability detect peak coefficient of rail surface realtime, locomotives are prone spinning and sliding phenomena under different conditions. This reduces utilization can even lead irreversible damage such as tread detachment wear. Therefore, this paper proposes a machine learning-based recognition method identify condition online set for surfaces. output motor...

Sponge city is an engineering solution proposed by the Chinese government which aims to deal with urban water issues (e.g., flooding, poor quality) brought on climate change and urbanization. Various strategies for sponge construction are required since environmental constraints differ regionally across country. To identify regional variations, reveal inner links between externalities design elements in construction, offer practical suggestions, efforts two directions made based data of 65...

In this report, we describe the technical details of our submission to 2024 RoboDrive Challenge Robust Map Segmentation Track. The track focuses on segmentation complex driving scene elements in BEV maps under varied conditions. Semantic map provides abundant and precise static environmental information crucial for autonomous systems' planning navigation. While current methods excel ideal circumstances, e.g., clear daytime conditions fully functional sensors, their resilience real-world...

CRYSTALS-Kyber is a new algorithm that the NIST recently selected to standardize public-key encryption and key establishment. Therefore, studies are needed evaluate side-channel attack resistance of implementations. This paper presents simple power analysis hardware implementation with security parameter k=3. Since implementations perform computations in parallel, consumption each operation difficult quantify. The entire trace was identified using 6,072,500 samples during CCAKEM Kyber. A...

With the increasing emphasis on social-awareness navigation of service robots, traditional social force models require improvements and recalibration. This paper proposes an enhanced model that integrates robot group forces with one. The is simulated in various scenes JRDB dataset, which categorized into three categories — outdoor, indoor&open, indoor&narrow. We use maximum likelihood estimation method genetic algorithm to calibrate parameters separately for each category provide explanation...

The automobile shock absorber is an important component of the automobile, and metal annular cylindrical workpiece indispensable part processing, so strict quality inspection for particularly important. There are many industrial products. Due to variety products complexity environment, it still difficult use a standardized method these numerous mechanical parts at this stage. For example, detection studied in paper manual end face workpiece, combines traditional image processing with depth...

In traditional recognition methods based on convolutional neural networks, the focus is improving accuracy and real-time issues, both of which depend large- scale datasets, but ignore issue unbalanced datasets. This extreme imbalance in sample set can cause model's performance to collapse. While transfer learning fine-tune existing pre-trained models using small-scale there still a significant challenge transitioning from (hundreds data) few samples. We use defect car shock absorber parts as...