We present a highly accurate and efficient, yet simple, two-stage voting scheme for distinguishing inlier 3D correspondences by densely assessing and ranking their local and global geometric consistencies. The strength of the proposed method stems from both the novel idea of post-validated voting set, as well as single-point superimposition transforms, which are computationally cheap and avoid rotational ambiguities. Using a well-known dataset consisting of various 3D models and numerous scenes that include different occlusion rates, the proposed scheme is evaluated against state-of-the-art 3D voting schemes, in terms of both the correspondence PR (precision-recall) AUC (area under curve), and the execution time. A total of 374 experiments were conducted for each method, which involved a combination of four models, 50 scenes, and two down-samplings. The proposed scheme outperforms the state-of-the-art 3D voting schemes in terms of both accuracy and speed. Quantitatively, the proposed scheme scores 97.0%±12.9% on the PR AUC metric, averaged over all of the experiments, while the two state-of-the-art schemes score 74.2%±22.2% and 78.3%±26.4%. Furthermore, the proposed scheme requires only 24.1%±6.0% of the time consumed by the fastest state-of-the-art scheme. The proposed voting scheme also demonstrates high robustness against occlusions and scarce inliers.

Artificial intelligence (AI) systems now challenge or surpass human experts in many computer games^1,2. Physical and real-time sports such as table tennis, however, remain a major open challenge because of their requirements for fast, precise and adversarial interactions near obstacles and at the edge of human reaction time³. Here we present Ace, to our knowledge the first real-world autonomous system competitive with elite human table tennis players. Ace addresses the challenges of physical real-time interaction through a new, high-speed perception system using event-based vision sensors⁴, and a new control system based on model-free reinforcement learning, as well as state-of-the-art high-speed robot hardware. Evaluated in matches against elite and professional players under official competition rules, Ace achieved several victories and demonstrated consistent returns of high-speed, high-spin shots. These results highlight the potential of physical AI agents to perform complex, real-time interactive tasks, suggesting broader applications in domains requiring fast, precise human-robot interaction.

Local image features are invariant to in-plane rotations and robust to minor viewpoint changes. However, the current detectors and descriptors for local image features fail to accommodate out-of-plane rotations larger than 25°–30°. Invariance to such viewpoint changes is essential for numerous applications, including wide baseline matching, 6D pose estimation, and object reconstruction. In this study, we present a general embedding that wraps a detector/descriptor pair in order to increase viewpoint invariance by exploiting input depth maps. The proposed embedding locates smooth surfaces within the input RGB-D images and projects them into a viewpoint invariant representation, enabling the detection and description of more viewpoint invariant features. Our embedding can be utilized with different combinations of descriptor/detector pairs, according to the desired application. Using synthetic and real-world objects, we evaluated the viewpoint invariance of various detectors and descriptors, for both standalone and embedded approaches. While standalone local image features fail to accommodate average viewpoint changes beyond 33.3°, our proposed embedding boosted the viewpoint invariance to different levels, depending on the scene geometry. Objects with distinct surface discontinuities were on average invariant up to 52.8°, and the overall average for all evaluated datasets was 45.4°. Similarly, out of a total of 140 combinations involving 20 local image features and various objects with distinct surface discontinuities, only a single standalone local image feature exceeded the goal of 60° viewpoint difference in just two combinations, as compared with 19 different local image features succeeding in 73 combinations when wrapped in the proposed embedding. Furthermore, the proposed approach operates robustly in the presence of input depth noise, even that of low-cost commodity depth sensors, and well beyond.

[abstFig src='/00280002/15.jpg' width=""300"" text='Teach grasping point by custom-made tool' ]We propose a method for conveying human knowledge to home and office assistance robots by teaching them how to perform the process of grasping objects with a custom-made tool. Specifically, we propose a semi-direct teaching method that respects the limitations of the hardware on the robot while utilizing human experience for intuitive teaching. We specify the information necessary for grasping objects through the generation of teaching data, which include the grasping force, relative position, and orientation. To respect the hardware limitations and at the same time allow inexperienced users to perform the teaching process easily, we used a teaching tool that possesses the same mechanism as the end effector of the robot. To simplify the teaching, we developed a sensing system that would reduce the teaching time with accurate measurements. Subsequently, the robot would use the teaching data to grasp the object. Experiments conducted using volunteers demonstrated the validity of the proposed method, wherein the teaching data for three different tasks were generated in less than 30 s each and accurate measurements were obtained for both the grasping position and force for grasping the objects.

Title: Refining Two Robots Task Execution Through Tuning Behavior Trajectory and Balancing the Communication | Keywords: cooperative manipulation, path refinement, human-robot interaction | Author: Jorge David Figueroa Heredia, Shouhei Shirafuji, Hamdi M. Sahloul, Jose Ildefonso U. Rubrico, Taiki Ogata, Tatsunori Hara, and Jun Ota

Foreground segmentation enables dynamic reconstruction of the moving objects in static scenes. After KinectFusion had proposed a novel method that constructs the foreground from the Iterative Closest Point (ICP) outliers, numerous studies proposed filtration methods to reduce outlier noise. To this end, the relationship between outliers and the foreground is investigated, and a method to efficiently extract the foreground from outliers is proposed. The foreground is found to be directly connected to ICP distance outliers rather than the angle and distance outliers that have been used in past research. Quantitative results show that the proposed method outperforms prevalent foreground extraction methods, and attains an average increase of 11.8% in foreground quality. Moreover, real-time speed of 50 fps is achieved without heavy graph-based refinements, such as GrabCut. The proposed depth features surpass current 3D GrabCut, which only uses RGB-N.