No abstract

No abstract

No abstract

As autonomous robots interact and navigate around real-world environments such as homes, it is useful to reliably identify and manipulate articulated objects, such as doors and cabinets. Many prior works in object articulation identification require manipulation of the object, either by the robot or a human. While recent works have addressed predicting articulation types from visual observations alone, they often assume prior knowledge of category-level kinematic motion models or sequence of observations where the articulated parts are moving according to their kinematic constraints. In this work, we propose training a neural network through largescale domain randomization to identify the articulation type of object parts from a single image observation. Training data is generated via photorealistic rendering in simulation. Our proposed model predicts motion residual flows of object parts, and these residuals are used to determine the articulation type and parameters. We train the network on six object categories with 149 objects and 100K rendered images, achieving an accuracy of 82.5%. Experiments show our method generalizes to novel object categories in simulation and can be applied to real-world images without fine-tuning.1 https://sim2realai.github.io/ Synthetic-Datasets-of-Objects-Part-I/

Background and PurposeResearch on stem cells (SC) is growing rapidly in neurology, but clinical applications of SC for neurological disorders remain to be proven effective and safe. Human clinical trials need to be registered in registries in order to reduce publication bias and selective reporting.MethodsWe searched three databases—clinicaltrials.gov, the Clinical Research Information System (CRIS), and PubMed—for neurologically relevant SC-based human trials and articles in Korea. The registration of trials, posting and publication of results, and registration of published SC articles were examined.ResultsThere were 17 completed trials registered at clinicaltrials.gov and the CRIS website, with results articles having been published for 5 of them. Our study found 16 publications, of which 1 was a review article, 1 was a protocol article, and 8 contained registered trial information.ConclusionsMany registered SC trials related to neurological disorders are not reported, while many SC-related publications are not registered in a public registry. These results support the presence of biased reporting and publication bias in SC trials related to neurological disorders in Korea.