In this paper, the Efficient Channel Attention (ECA) mechanism is incorporated at terminal layer of YOLOv10 backbone network to enhance feature expression capability. addition, Transformer introduced into C3 module in extraction process construct C3TR replace original C2F as deepening module. study, both ECA and self-attention are thoroughly analyzed integrated YOLOv10. The used an important part deepen effect extraction. model long-distance dependency relationship, capture global contextual...

The traffic sign detection algorithm based on Faster Region-Based Convolutional Neural Network (R-CNN) has been applied to various intelligent-vehicles driving scenarios. However, the model of current certain shortcomings, which include influence weather and light, distance signs, similar signs. To solve these problems, this paper proposes an improved method R-CNN. First, we propose a fusion that fuses feature pyramid into R-CNN algorithm. This can extract object features with precision...

This paper represents a constraint planning and optimization control scheme for highly redundant mobile manipulator considering complex indoor environment. Compared with the traditional solution of manipulator, infinity norm slack variable are additionally introduced leveraged by algorithm. The former takes into account joint limits effectively individual velocities latter relaxes equality decreasing infeasible area. By using derived kinematic equations, tracking problem is expressed as an...

To solve the component segmentation problem caused by sticking and overlapping of parts in incoherent handwritten calligraphy characters, we propose a Chinese character part method based on Faster RCNN. The utilizes advantages RCNN multi-scale small targets to difficult problems segmentation. hierarchical features components were used our proposed identify each layer structure obtain components. Qualitative quantitative calculations test effect method. experimental results demonstrate...

Data persistence is necessary for many in-memory applications. However, the disk-based data largely slows down Emerging non-volatile memory (NVM) offers an opportunity to achieve at DRAM-level performance. Nevertheless, NVM typically requires a software library operate data, which brings significant overhead. This article demonstrates that hardware-based high-frequency checkpointing mechanism can be used efficient on NVM. To maintain checkpoint consistency, traditional logging and...

Nowadays emerging Non-Volatile Memory (NVM) technologies are introduced to remedy the shortages of current DRAM-based memory system. However, NVM has limited write endurance, which would severely restrict performance In order relieve above limitation, we proposed a victim-aware cache policy (VAIL), based on DRAM and hybrid VAIL took eviction locality victims from into consideration, reduce writebacks improve hit ratio at same time. Our evaluation shows that could 17.2% with single-core...

We propose a calibration method for automotive augmented reality head-up displays (AR-HUD) using chessboard pattern and warping maps. The HUD is modeled as pinhole camera whose intrinsic parameters are determined by employing stereo method. select several viewpoints within the driver's eye box place smartphone at each of them in sequence, position sensed head tracker. By automatically shifting 2D points on virtual image to 3D corners view camera, we obtain group 2D–3D correspondences then...

In this paper we create "three-point service-oriented conceptual design and modeling methodology" for Web services composition based on the object-oriented research results in MERODE, thus concept level firstly clarify issue semantic consistency of composition. Afterwards, logic precisely define demonstrate through using some standard schemes an order system. These aspects can effectively significantly tackle one big inhibitors adoption: a lack business processes within interactions

In this paper, we introduce a newly developed target-free calibration method for automotive augmented reality head-up displays (AR-HUDs), which can be performed fully automatically using smartphone camera. Our requires no target to set up in front of the vehicle. Instead, it utilizes feature points environment, makes robust against misplaced targets and allows an easy deployment, i.e. garages. Under pinhole model assumption, decouple perspective projection matrix into three parts: intrinsic...

Considering the limited communications conditions such as delays, disturbances, and topologies uncertainties, stability criteria for robust consensus of multiagent systems are proposed in this paper. Firstly, by using idea state decomposition space transformation, condition guaranteeing is converted into verifying disagreement system. In order to deal with multiple time-varying delays switching topologies, jointly quadratic common Lyapunov-Krasovskii (JQCLK) functional built analyze...

Since Chinese characters are composed of components, deforming the components in a small number existing calligraphy to generate new is an effective method produce character library same style. Usually, component deformation achieved by affine transformation. However, when calculating parameters transformation, methods usually have problems large amount manual participation or complicated calculation. In this paper, we proposed Analytic Hierarchy Process (AHP)-based method, which simple...