Learning by Observation without Three-Dimensional Reconstruction
نویسندگان
چکیده
Abstract. This paper presents a basic skill of imitation learning from observation towards developmental cognitive robotics which aims at providing a new way to understand the cognitive process of human beings by realizing artificial systems based on the internal observer’s viewpoint. Unlike the conventional ones, the proposed method does not assume the global coordinate system in order to transform the observed motion in the demonstrator’s coordinate system to the learner’s one. Instead, the optic-geometrical constraint called “epipolar constraint” is used to reconstruct the view of the other agent, on which adaptive visual servoing is applied to imitate the observed motion. The experimental results and future issues are given.
منابع مشابه
Three-dimensional reconstruction of New Zealand rabbit antebrachium by multidetector computed tomography
The aim of this study was to reveal biometric peculiarities of New Zealand white rabbit antebrachium (radius and ulna) by means of three-dimensional (3D) reconstruction of multidetector computed tomography (MDCT) images. Under general anesthesia, the antebrachiums of a total of sixteen rabbits of both sexes were scanned with a general diagnostic MDCT. Biometric measurements of the reconstructed...
متن کاملFetus and uteroplacental unit: fast MR imaging with three-dimensional reconstruction and volumetry--feasibility study.
In 22 fetuses, three-dimensional reconstruction was performed on the basis of T2-weighted magnetic resonance (MR) data sets by using a surface-rendering technique. Fetuses and uteroplacental units were segmented manually; volumes were calculated automatically. Three-dimensional reconstruction and volumetry of the fetus and uteroplacental unit are feasible with MR data sets. MR estimates correla...
متن کاملPerspective Transformer Nets: Learning Single-View 3D Object Reconstruction without 3D Supervision
Understanding the 3D world is a fundamental problem in computer vision. However, learning a good representation of 3D objects is still an open problem due to the high dimensionality of the data and many factors of variation involved. In this work, we investigate the task of single-view 3D object reconstruction from a learning agent’s perspective. We formulate the learning process as an interact...
متن کاملTarget detection The Simulation Analysis of the Effect of Imaging Sidelap/Overlap on the Quality of 3D Reconstruction in UAV Photogrammetry and Determination of their Optimal Values
The selection of effective and optimal sidelap/overlap of images according to the quality, speed, cost and purpose of each UAV photogrammetric project is considered as an important parameter. Since two decades ago, several researches have been carried out based on a trial and error strategy for this purpose, in which different drones and cameras have been utilized for imaging with different sid...
متن کاملEffect of three-dimensional reconstruction-assisted 23G micro-invasive vitrectomy in patients with proliferative diabetic retinopathy
In the present study, we investigated the effect of optical coherence tomography (OCT) three-dimensional reconstruction-assisted 23G micro-invasive vitrectomy (abbreviated to'23G') in patients with proliferative diabetic retinopathy (PDR). A total of 66 PDR patients (66 eyes) were continuously selected and randomly divided into the control and observation groups with 33 patients in each group. ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2004