The Experts below are selected from a list of 2682 Experts worldwide ranked by ideXlab platform
Alexander Reshetov - One of the best experts on this subject based on the ideXlab platform.
-
faster ray packets triangle intersection through vertex culling
2007 IEEE Symposium on Interactive Ray Tracing, 2007Co-Authors: Alexander ReshetovAbstract:Acceleration structures are used in ray tracing to sharply reduce number of ray-triangle intersection tests at the expense of traversing such structures. Bigger structures eliminate more tests, but their traversal becomes less efficient, especially for ray packets, for which number of inactive rays increases at the lower levels of the acceleration structures. For dynamic scenes, building or updating acceleration structures is one of the major performance impediments. We propose a new way to reduce the total number of tests by creating a special transient Frustum every time a leaf is traversed by a packet of rays. This Frustum contains intersections of active rays with a leaf node and eliminates over 90% of all potential tests. It allows a tenfold reduction in size of acceleration structure whilst still achieving a better performance.
-
faster ray packets triangle intersection through vertex culling
International Conference on Computer Graphics and Interactive Techniques, 2007Co-Authors: Alexander ReshetovAbstract:To eliminate unnecessary ray packet-triangle intersection tests, we check for separation of two convex objects: a triangle and a Frustum containing intersections of rays with a leaf node of an acceleration structure. We show a performance improvement that is proportional to the ratio of lengths of average leaf edge to triangle edge, which opens new possibilities for creation of better acceleration structures.
Kris M Kitani - One of the best experts on this subject based on the ideXlab platform.
-
monocular 3d object detection with pseudo lidar point cloud
International Conference on Computer Vision, 2019Co-Authors: Xinshuo Weng, Kris M KitaniAbstract:Monocular 3D scene understanding tasks, such as object size estimation, heading angle estimation and 3D localization, is challenging. Successful modern-day methods for 3D scene understanding require the use of a 3D sensor. On the other hand, single image-based methods have significantly worse performance. In this work, we aim at bridging the performance gap between 3D sensing and 2D sensing for 3D object detection by enhancing LiDAR-based algorithms to work with single image input. Specifically, we perform monocular depth estimation and lift the input image to a point cloud representation, which we call pseudo-LiDAR point cloud. Then we can train a LiDAR-based 3D detection network with our pseudo-LiDAR end-to-end. Following the pipeline of two-stage 3D detection algorithms, we detect 2D object proposals in the input image and extract a point cloud Frustum from the pseudo-LiDAR for each proposal. Then an oriented 3D bounding box is detected for each Frustum. To handle the large amount of noise in the pseudo-LiDAR, we propose two innovations: (1) use a 2D-3D bounding box consistency constraint, adjusting the predicted 3D bounding box to have a high overlap with its corresponding 2D proposal after projecting onto the image; (2) use the instance mask instead of the bounding box as the representation of 2D proposals, in order to reduce the number of points not belonging to the object in the point cloud Frustum. Through our evaluation on the KITTI benchmark, we achieve the top-ranked performance on both bird's eye view and 3D object detection among all monocular methods, effectively quadrupling the performance over previous state-of-the-art. Our code is available at https://github.com/xinshuoweng/Mono3D_PLiDAR.
-
monocular 3d object detection with pseudo lidar point cloud
arXiv: Computer Vision and Pattern Recognition, 2019Co-Authors: Xinshuo Weng, Kris M KitaniAbstract:Monocular 3D scene understanding tasks, such as object size estimation, heading angle estimation and 3D localization, is challenging. Successful modern day methods for 3D scene understanding require the use of a 3D sensor. On the other hand, single image based methods have significantly worse performance. In this work, we aim at bridging the performance gap between 3D sensing and 2D sensing for 3D object detection by enhancing LiDAR-based algorithms to work with single image input. Specifically, we perform monocular depth estimation and lift the input image to a point cloud representation, which we call pseudo-LiDAR point cloud. Then we can train a LiDAR-based 3D detection network with our pseudo-LiDAR end-to-end. Following the pipeline of two-stage 3D detection algorithms, we detect 2D object proposals in the input image and extract a point cloud Frustum from the pseudo-LiDAR for each proposal. Then an oriented 3D bounding box is detected for each Frustum. To handle the large amount of noise in the pseudo-LiDAR, we propose two innovations: (1) use a 2D-3D bounding box consistency constraint, adjusting the predicted 3D bounding box to have a high overlap with its corresponding 2D proposal after projecting onto the image; (2) use the instance mask instead of the bounding box as the representation of 2D proposals, in order to reduce the number of points not belonging to the object in the point cloud Frustum. Through our evaluation on the KITTI benchmark, we achieve the top-ranked performance on both bird's eye view and 3D object detection among all monocular methods, effectively quadrupling the performance over previous state-of-the-art. Our code is available at this https URL.
Hamed Jamshidi Aval - One of the best experts on this subject based on the ideXlab platform.
-
Influences of pin profile on the mechanical and microstructural behaviors in dissimilar friction stir welded AA6082-AA7075 butt Joint
Materials and Design, 2015Co-Authors: Hamed Jamshidi AvalAbstract:This paper investigates the effect of tool pin profile and post weld heat treatment on the mechanical and microstructural aspects of friction stir welded dissimilar precipitation-hardenable aluminum alloys between 6xxx (Al-Mg-Si) and 7xxx (Al-Zn-Mg). It was found that the tool with a conical probe with three grooves produces higher temperatures than the tool with a square Frustum probe. The mixture of materials in the weld nugget of joints welded by the tool with the square Frustum probe was more uniform. Of the natural aging times studied, the stir zone of all welds showed the highest kinetics of microhardness and strength recovery; its effect increased considerably in the heat input per unit length decreased.
Noriyuki Kato - One of the best experts on this subject based on the ideXlab platform.
-
analysis of circular trajectory equivalent to cone Frustum milling in five axis machining centers using motion simulator
International Journal of Machine Tools & Manufacture, 2013Co-Authors: Noriyuki Kato, Masaomi Tsutsumi, Ryuta SatoAbstract:Abstract The present paper describes the effect of the half apex angle of the cone-Frustum on the motion trajectory under simultaneous five-axis motion and the effect of the sensitive direction of the ball bar when the motion trajectory is measured along the three-dimensional circular conical path. In the present paper, simulation of the measurement by means of a ball bar instrument is mainly conducted using a motion simulator developed previously. In particular, a precise mathematical model was developed to express the pitch errors of the axes of rotation of the five-axis machining center having a tilting rotary table driven by worm gears. In the experiment and simulation, primarily the center position and half apex angle of the cone-Frustum were varied. In addition, two sensitive directions of the ball bar were investigated. The motion simulator incorporating the pitch error model can express the detailed trajectories obtained by the ball bar, even if the half apex angle and center position of the cone-Frustum and the sensitive direction of the ball bar were changed. Then, the influence of the frictional force of the linear axes of motion, and the backlash and pitch error of the axes of rotation on the circular trajectories were analyzed. In particular, for the case of a half apex angle of 45°, the trajectory due to the errors of the axis of rotation is strongly affected by the sensitive direction of the ball bar.
-
3d circular interpolation motion equivalent to cone Frustum cutting in five axis machining centers and its sensitivity analysis
Procedia CIRP, 2012Co-Authors: Noriyuki Kato, Ryuta Sato, Masaomi TsutsumiAbstract:Abstract The present paper describes a testing method for five-axis machining centers using three-dimensional circular interpolation movement equivalent to cone-Frustum cutting. In the present paper, the test conditions, such as the half apex angle of cone-Frustum and the sensitive directions of the ball bar device were investigated. In addition, the sensitivity coefficient of each axis was investigated. It is found from the analysis of the sensitivity coefficient that the trajectory due to the errors of the axis of rotation is strongly affected by the sensitive direction of the ball bar for the case of a half apex angle of 45°
-
analysis of nc data for machining test of cone Frustum under simultaneous five axis control
Transactions of the Japan Society of Mechanical Engineers. C, 2011Co-Authors: Noriyuki Kato, Daisuke Yumiza, Masaomi Tsutsumi, Yu Tsuchihashi, Chengri Cui, Yukitoshi IharaAbstract:NC data for machining the cone-Frustum by means of five-axis machining centers with a tilting rotary table are analyzed in detail. The cone-Frustum is machined by using a milling cutter at simultaneous five-axis control for inspecting the machining accuracy. However, the NC data have to be changed corresponding to the setting position of the workpiece on the rotary table. Thus, the moving range and velocity of each axis are analyzed by changing the apex angle and center position of the cone. It is found from the analyzed results that the larger apex angle gives the larger moving range and the higher velocity to all axes. When the center position moves along the Y-direction, every reversal point independently appears on a circular path. Therefore, it is recommended to offset the center position of the cone Frustum along Y-direction to give the large moving ranges and the high velocities to the axes.
Ryuta Sato - One of the best experts on this subject based on the ideXlab platform.
-
analysis of circular trajectory equivalent to cone Frustum milling in five axis machining centers using motion simulator
International Journal of Machine Tools & Manufacture, 2013Co-Authors: Noriyuki Kato, Masaomi Tsutsumi, Ryuta SatoAbstract:Abstract The present paper describes the effect of the half apex angle of the cone-Frustum on the motion trajectory under simultaneous five-axis motion and the effect of the sensitive direction of the ball bar when the motion trajectory is measured along the three-dimensional circular conical path. In the present paper, simulation of the measurement by means of a ball bar instrument is mainly conducted using a motion simulator developed previously. In particular, a precise mathematical model was developed to express the pitch errors of the axes of rotation of the five-axis machining center having a tilting rotary table driven by worm gears. In the experiment and simulation, primarily the center position and half apex angle of the cone-Frustum were varied. In addition, two sensitive directions of the ball bar were investigated. The motion simulator incorporating the pitch error model can express the detailed trajectories obtained by the ball bar, even if the half apex angle and center position of the cone-Frustum and the sensitive direction of the ball bar were changed. Then, the influence of the frictional force of the linear axes of motion, and the backlash and pitch error of the axes of rotation on the circular trajectories were analyzed. In particular, for the case of a half apex angle of 45°, the trajectory due to the errors of the axis of rotation is strongly affected by the sensitive direction of the ball bar.
-
3d circular interpolation motion equivalent to cone Frustum cutting in five axis machining centers and its sensitivity analysis
Procedia CIRP, 2012Co-Authors: Noriyuki Kato, Ryuta Sato, Masaomi TsutsumiAbstract:Abstract The present paper describes a testing method for five-axis machining centers using three-dimensional circular interpolation movement equivalent to cone-Frustum cutting. In the present paper, the test conditions, such as the half apex angle of cone-Frustum and the sensitive directions of the ball bar device were investigated. In addition, the sensitivity coefficient of each axis was investigated. It is found from the analysis of the sensitivity coefficient that the trajectory due to the errors of the axis of rotation is strongly affected by the sensitive direction of the ball bar for the case of a half apex angle of 45°
