Automated tools which can understand and interface with CAD (computer-aided design) models are of significant research interest due to the potential for improving efficiency in manufacturing processes. At present, most research into the use of artificial intelligence to interpret three-dimensional data takes input in the form of multiple two-dimensional images of the object or in the form of three-dimensional grids of voxels. The transformation of the input data necessary for these approaches inevitably leads to some loss of information and limitations of resolution. Existing research into the direct analysis of model files in STEP (standard for the exchange of product data) format tends to follow a rules-based approach to analyse models of a certain type, resulting in algorithms without the benefits of flexibility and complex understanding which artificial intelligence can provide. In this paper, a novel recursive encoder network for the automatic analysis of STEP files is presented. The encoder network is a flexible model with the potential for adaptation to a wide range of tasks and finetuning for specific CAD model datasets. Performance is evaluated using a machining feature classification task, with results showing accuracy approaching 100% and training time comparable to that of existing multi-view and voxel-based solutions without the need for a GPU.
This paper outlines the development of a non-intrusive alternative to current intelligent transportation systems using road-side video cameras. The use of video to determine the axle count and speed of vehicles traveling on major roads was investigated. Two instances of a convolutional neural network, YOLOv3, were trained to perform object detection for the purposes of axle detection and speed measurement, achieving accuracies of 95% and 98% mAP respectively. Outputs from the axle detection were processed to produce axle counts for each vehicle with 93% accuracy across all vehicles where all axles are visible. A simple Kalman filter was used to track the vehicles across the video frame, which worked well but struggled with longer periods of occlusion. The camera was calibrated for speed measurement using road markings in place of a reference object. The calibration method proved to be accurate, however, a constant error was introduced if the road markings were not consistent with the government specifications. The average vehicle speeds calculated were within the expected range. Both models achieved real-time speed performance.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.