The proceedings contain 31 papers. The special focus in this conference is on Precision Assembly Seminar. The topics include: High speed and low weight micro actuators for high precision assembly applications: Micro A...
ISBN:
(纸本)9780387312767
The proceedings contain 31 papers. The special focus in this conference is on Precision Assembly Seminar. The topics include: High speed and low weight micro actuators for high precision assembly applications: Micro Actuators with the Micro Harmonic Drive®;automated assembly planning based on skeleton modelling strategy;morphological classification of hybrid microsystems assembly;First steps in integrating micro-assembly features into industrially used DFA software;tolerance budgeting in a novel coarse-fine strategy for micro-assembly;the importance of concept and design visualisation in the production of an automated assembly and test machine;development of passive alignment techniques for the assembly of hybrid microsystems;miniature reconfigurable assembly line for small products;multi-axes micro gripper for the handling and alignment of flexible micro parts: Development of compact and shock resistant gripper components;conception of a scalable production for micro-mechatronical products: Systematics for planning and platform with process modules for the production of micro-mechatronicalproducts;towards an integrated assembly process decomposition and modular equipment configuration: A knowledge enhanced iterative approach;evolvable skills for assembly systems: Evolvability by automatic configuration and standardization of control interfaces and state models;toward the vision based supervision of microfactories through images mosaicing;precision multi-degrees-of-freedom positioning systems: Modular design for assembly applications;what is the best way to increase efficiency in precision assembly?;life cycle and cost analysis for modular re-configurable final assembly systems;impact of bad components on costs and productivity in automatic assembly;the reliable application of average and highly viscous media: Assembly Net;laser sealed packaging for microsystems;preface.
The practice of yoga encompasses mental, physical, and spiritual dimensions, aiming for holistic wellbeing. Accurate alignment in yoga enhances the effectiveness of each pose by targeting specific muscle groups, reduc...
详细信息
ISBN:
(数字)9798350375190
ISBN:
(纸本)9798350375206
The practice of yoga encompasses mental, physical, and spiritual dimensions, aiming for holistic wellbeing. Accurate alignment in yoga enhances the effectiveness of each pose by targeting specific muscle groups, reducing strain on muscles and joints, and improving stability and balance. This research employs advanced computer visiontechniques, YOLO (You Only Look Once) and MediaPipe to identify critical keypoints from the skeletal structures of yoga practitioners, thereby providing a detailed representation of body alignment for posture recognition. Augmented using the SMOTE technique, the skeletal data serves as input for various machine Learning and ensemble models during the training process. The study utilizes a 2D image dataset comprising 20 well-known yoga poses. Among the models tested, the LightGBM ensemble classifier using MediaPipe keypoints achieved the highest accuracy at 96.52%. Further analysis included the evaluation of the model through a confusion matrix, learning curve, and pose-wise accuracy, even for similar-looking exercises. These findings highlight the potential of integrating computer vision and machine learning to enhance yoga practice through precise posture recognition and alignment analysis.
One of the main roles played by real time image segmentation is to enhance and catalyse self driving cars that can accurately sense their surroundings due to in terms of proper functioning. The new model was proposed ...
详细信息
ISBN:
(数字)9798350375190
ISBN:
(纸本)9798350375206
One of the main roles played by real time image segmentation is to enhance and catalyse self driving cars that can accurately sense their surroundings due to in terms of proper functioning. The new model was proposed with a completely new image segmentation paradigm for enabling self-driving in real time. The deployment of deep learning techniques, accurate real-time object segmentation through a fusion of fully convolutional network (FCN) with multilevel pyramid analysis. FCN uses the input image for segmentation, and in return creates a map-based on pixel-wise classification from which we can get more detailed information about different objects present in the scene. This approach creates a very large FCN network that is combined with a multiscale pyramid architecture that process an image in multiple stages effectively collecting complex surroundings information to improve performance. The proposed method is then experimented on different data with a particular focus on challenging to adopt such infrastructure in order to achieve modern day results in terms of accuracy and real time performance.
暂无评论