Tool manipulation is vital for facilitating robots to complete challenging task goals. It requires reasoning about the desired effect of the task and thus properly grasping and manipulating the tool to achieve the task. Task-agnostic grasping optimizes for grasp robustness while ignoring crucial taskspecific constraints. In this paper, we propose the Task-Oriented Grasping Network (TOG-Net) to jointly optimize both taskoriented grasping of a tool and the manipulation policy for that tool. The training process of the model is based on largescale simulated self-supervision with procedurally generated tool objects. We perform both simulated and real-world experiments on two tool-based manipulation tasks: sweeping and hammering. Our model achieves overall 71.1% task success rate for sweeping and 80.0% task success rate for hammering. Supplementary material is available at: bit.ly/task-oriented-grasp.
3D reconstruction from a single image is a key problem in multiple applications ranging from robotic manipulation to augmented reality. Prior methods have tackled this problem through generative models which predict 3D reconstructions as voxels or point clouds. However, these methods can be computationally expensive and miss fine details. We introduce a new differentiable layer for 3D data deformation and use it in DEFORMNET to learn a model for 3D reconstructionthrough-deformation. DEFORMNET takes an image input, searches the nearest shape template from a database, and deforms the template to match the query image. We evaluate our approach on the ShapeNet dataset and show that -(a) the Free-Form Deformation layer is a powerful new building block for Deep Learning models that manipulate 3D data (b) DEFORMNET uses this FFD layer combined with shape retrieval for smooth and detail-preserving 3D reconstruction of qualitatively plausible point clouds with respect to a single query image (c) compared to other state-of-the-art 3D reconstruction methods, DEFORMNET quantitatively matches or outperforms their benchmarks by significant margins. For more information, visit: https://deformnet-site.github.io/DeformNet-website/.
A low-cost resource approach to ADHD therapy would be a practical approach to treating children in developing countries. Research has shown that ADHD is prevalent in all areas of the world, and yet treatment for children in more impoverished countries is still lacking. The approach taken was to combine yoga and meditation combined with multimodal behavioral therapy program for children ageing 6 to 11. The program was kept low cost by using trained high school volunteers and integrating the program within the public school. After 6 weeks of the program, 90.5% of children showed improvement as measured by their performance impairment score, a measurement of academic performance. Parent and Teacher evaluations of behavior also found improvement as 25 of the 64 children (39.1%) improved into the normal range as measured by the Vanderbilt questionnaire. Moreover, children could successfully learn both yoga and meditation from high school students irrespective of their age, ADHD type, or initial performance impairment. The results demonstrate efficacy of a multimodal behavioral program incorporating yoga and meditation. The use of high school volunteers from schools in the area demonstrates an effective low-cost and universally applicable approach.
The objective was to assess the efficacy of a one-year, peer-mediated interventional program consisting of yoga, meditation and play therapy maintained by student volunteers in a school in India. The population consisted of 69 students between the ages of 6 and 11 years, previously identified as having attention deficit hyperactivity disorder (ADHD). A program, known as Climb-Up, was initially embedded in the school twice weekly. Local high school student volunteers were then trained to continue to implement the program weekly over the period of one year. Improvements in ADHD symptoms and academic performance were assessed using Vanderbilt questionnaires completed by both parents and teachers. The performance impairment scores for ADHD students assessed by teachers improved by 6 weeks and were sustained through 12 months in 46 (85%) of the enrolled students. The improvements in their Vanderbilt scores assessed by parents were also seen in 92% (P < 0.0001, Wilcoxon). The Climb-Up program resulted in remarkable improvements in the students' school performances that were sustained throughout the year. These results show promise for a cost-effective program that could easily be implemented in any school.
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.