scholarly journals Adversarial discriminative sim-to-real transfer of visuo-motor policies

2019 ◽  
Vol 38 (10-11) ◽  
pp. 1229-1245 ◽  
Author(s):  
Fangyi Zhang ◽  
Jürgen Leitner ◽  
Zongyuan Ge ◽  
Michael Milford ◽  
Peter Corke
Keyword(s):  
Success Rate ◽  
Real World ◽  
Real Data ◽  
Moving Target ◽  
Reaching Task ◽  
Data Requirement ◽  
Modular Networks ◽  
Robotic Applications ◽  
Velocity Mode ◽  
Control Accuracy

Various approaches have been proposed to learn visuo-motor policies for real-world robotic applications. One solution is first learning in simulation then transferring to the real world. In the transfer, most existing approaches need real-world images with labels. However, the labeling process is often expensive or even impractical in many robotic applications. In this article, we introduce an adversarial discriminative sim-to-real transfer approach to reduce the amount of labeled real data required. The effectiveness of the approach is demonstrated with modular networks in a table-top object-reaching task where a seven-degree-of-freedom arm is controlled in velocity mode to reach a blue cuboid in clutter through visual observations from a monocular RGB camera. The adversarial transfer approach reduced the labeled real data requirement by 50%. Policies can be transferred to real environments with only 93 labeled and 186 unlabeled real images. The transferred visuo-motor policies are robust to novel (not seen in training) objects in clutter and even a moving target, achieving a 97.8% success rate and 1.8 cm control accuracy. Datasets and code are openly available.

scholarly journals Improving the performance of a radio-frequency localization system in adverse outdoor applications

2021 ◽  
Vol 2021 (1) ◽  
Author(s):  
Marcelo N. de Sousa ◽  
Ricardo Sant’Ana ◽  
Rigel P. Fernandes ◽  
Julio Cesar Duarte ◽  
José A. Apolinário ◽  
...  
Keyword(s):  
Random Forest ◽  
Ray Tracing ◽  
Real World ◽  
Practical Implication ◽  
Real Life ◽  
Simulated Data ◽  
Real Data ◽  
Gradient Boosting ◽  
Real World Data ◽  
Localization Accuracy

AbstractIn outdoor RF localization systems, particularly where line of sight can not be guaranteed or where multipath effects are severe, information about the terrain may improve the position estimate’s performance. Given the difficulties in obtaining real data, a ray-tracing fingerprint is a viable option. Nevertheless, although presenting good simulation results, the performance of systems trained with simulated features only suffer degradation when employed to process real-life data. This work intends to improve the localization accuracy when using ray-tracing fingerprints and a few field data obtained from an adverse environment where a large number of measurements is not an option. We employ a machine learning (ML) algorithm to explore the multipath information. We selected algorithms random forest and gradient boosting; both considered efficient tools in the literature. In a strict simulation scenario (simulated data for training, validating, and testing), we obtained the same good results found in the literature (error around 2 m). In a real-world system (simulated data for training, real data for validating and testing), both ML algorithms resulted in a mean positioning error around 100 ,m. We have also obtained experimental results for noisy (artificially added Gaussian noise) and mismatched (with a null subset of) features. From the simulations carried out in this work, our study revealed that enhancing the ML model with a few real-world data improves localization’s overall performance. From the machine ML algorithms employed herein, we also observed that, under noisy conditions, the random forest algorithm achieved a slightly better result than the gradient boosting algorithm. However, they achieved similar results in a mismatch experiment. This work’s practical implication is that multipath information, once rejected in old localization techniques, now represents a significant source of information whenever we have prior knowledge to train the ML algorithm.

scholarly journals RON-Gauss: Enhancing Utility in Non-Interactive Private Data Release

2019 ◽  
Vol 2019 (1) ◽  
pp. 26-46 ◽  
Author(s):  
Thee Chanyaswad ◽  
Changchang Liu ◽  
Prateek Mittal
Keyword(s):  
Machine Learning ◽  
Real World ◽  
Differential Privacy ◽  
Real Data ◽  
The Novel ◽  
Private Data ◽  
Data Release ◽  
Machine Learning Applications ◽  
Order Of Magnitude ◽  
Real World Datasets

Abstract A key challenge facing the design of differential privacy in the non-interactive setting is to maintain the utility of the released data. To overcome this challenge, we utilize the Diaconis-Freedman-Meckes (DFM) effect, which states that most projections of high-dimensional data are nearly Gaussian. Hence, we propose the RON-Gauss model that leverages the novel combination of dimensionality reduction via random orthonormal (RON) projection and the Gaussian generative model for synthesizing differentially-private data. We analyze how RON-Gauss benefits from the DFM effect, and present multiple algorithms for a range of machine learning applications, including both unsupervised and supervised learning. Furthermore, we rigorously prove that (a) our algorithms satisfy the strong ɛ-differential privacy guarantee, and (b) RON projection can lower the level of perturbation required for differential privacy. Finally, we illustrate the effectiveness of RON-Gauss under three common machine learning applications – clustering, classification, and regression – on three large real-world datasets. Our empirical results show that (a) RON-Gauss outperforms previous approaches by up to an order of magnitude, and (b) loss in utility compared to the non-private real data is small. Thus, RON-Gauss can serve as a key enabler for real-world deployment of privacy-preserving data release.

scholarly journals Marketing Relevance Through Market Theory

2018 ◽  
Vol 17 (5) ◽  
pp. 730-746 ◽  
Author(s):  
Stephen L. Vargo
Keyword(s):  
Real World ◽  
Applied Research ◽  
Real Data ◽  
Market Theory ◽  
Academic Marketing ◽  
Real World Problems

There have been numerous calls for more relevance in academic marketing, both for and by practitioners and for customers (e.g., Sheth and Sisodia 2006, Hunt 2018, Jaworski, Kohli, and Sahay 2000). It might seem that these calls signal the need for more applied research, based on real data and real-world problems. However, it seems to me that there has never before been such a plethora of empirical articles in marketing journals as there are presently and thus the problem must be much more basic.

The failure of certain fractional calculus operators in two physical models

2019 ◽  
Vol 22 (2) ◽  
pp. 255-270 ◽  
Author(s):  
Manuel D. Ortigueira ◽  
Valeriy Martynyuk ◽  
Mykola Fedula ◽  
J. Tenreiro Machado
Keyword(s):  
Fractional Calculus ◽  
Human Body ◽  
Real World ◽  
Fractional Derivatives ◽  
Electrical Impedance ◽  
Real Data ◽  
Physical Models ◽  
Data Sets ◽  
Real World Data ◽  
Fractional Calculus Operators

Abstract The ability of the so-called Caputo-Fabrizio (CF) and Atangana-Baleanu (AB) operators to create suitable models for real data is tested with real world data. Two alternative models based on the CF and AB operators are assessed and compared with known models for data sets obtained from electrochemical capacitors and the human body electrical impedance. The results show that the CF and AB descriptions perform poorly when compared with the classical fractional derivatives.

VP07 Collaboratively Modelling The Impact Of Interventions Retrospectively

2017 ◽  
Vol 33 (S1) ◽  
pp. 149-149
Author(s):  
Gordon Bache ◽  
Sukh Tatla ◽  
Deborah Simpson
Keyword(s):  
Real World ◽  
Real Data ◽  
System Level ◽  
Local Data ◽  
Real World Data ◽  
Total Saving ◽  
Retrospective Assessment ◽  
Administration Time ◽  
The Individual ◽  
The Impact

INTRODUCTION:A conventional approach to communicating value is to model the budget impact of a medicine and the associated formulations in which it is available to be prescribed. However, such an approach does not demonstrate the actual realization of the proposed impact. This abstract outlines an approach to presenting retrospective data back to healthcare professionals (HCP) that blends assumptions and real-world data. For illustrative purposes, we present the results of an application of the model for subcutaneously delivered trastuzumab in an anonymized trust in Yorkshire and Humber.METHODS:The authors developed a model that examined one calendar year (from April 2014) of redistributed sales data for both the intravenous and subcutaneous formulations of trastuzumab for every National Health Service (NHS) trust in England. A series of baseline assumptions (1) were used to model the resource impact of different formulations such as chair time, HCP time, pharmacy preparation time, consumables, wastage, and other considerations. Impacts were estimated at the individual attendance level and scaled to the caseload. These baseline assumptions could then be overwritten by the individual trust using local data.RESULTS:The site delivered approximately 985 doses of subcutaneous trastuzumab over a period of 12 months from April 2014, which represented about 76 percent of the total number of doses delivered. Chair time is estimated to have reduced by 22 minutes per attendance, resulting in a total saving of 361hours. HCP administration time is estimated to have reduced by 23 minutes per attendance, resulting in a total saving of 378 hours based on changing 985 IV doses to SC therapy.CONCLUSIONS:Blending real data and assumptions to provide a retrospective assessment of actual benefits realized back to HCPs is a powerful tool for demonstrating real-world value at both an individual trust and system level.

scholarly journals Effectiveness and safety of daptomycin in patients with infective endocarditis undergoing heart valve replacement: a subgroup analysis from real-world data

2017 ◽  
Vol 4 (2) ◽  
pp. 41-47 ◽  
Author(s):  
Achyut Guleri ◽  
Riccardo Utili ◽  
Pascal Dohmen ◽  
Kamal Hamed
Keyword(s):  
Infective Endocarditis ◽  
Success Rate ◽  
Clinical Outcomes ◽  
Valve Replacement ◽  
Heart Valve ◽  
Real World ◽  
Clinical Success ◽  
Clinical Success Rate ◽  
Heart Valve Replacement ◽  
Real World Data

Background: European Cubicin® Outcomes Registry and Experience (EU-CORE) was a retrospective, non-interventional, multicenter registry that collected real-world clinical outcomes following daptomycin use for the treatment of Gram-positive infections. EU-CORE data from patients with infective endocarditis (IE) who underwent heart valve replacement were analysed. Methods: Clinical outcomes were assessed as success (cured or improved), failure, or non-evaluable. Adverse events (AEs) were recorded for up to 30 days after daptomycin treatment. Results: Of 610 patients with IE, 198 [32.5%; left-sided IE (LIE), 166 (83.8%); right-sided IE (RIE), 21 (10.6%); both LIE and RIE, 11 (5.6%)] underwent heart valve replacement. Other than cardiovascular disease, renal disease (18.2%), sepsis (16.2%), and diabetes mellitus (15.2%) were the most significant underlying diseases. Major pathogens in patients with positive culture results (68.0%) were Staphylococcus aureus [36.8%; methicillin-resistant S. aureus (MRSA), 12.8%] and coagulase-negative staphylococci (CoNS; 31.6%). Daptomycin treatment [median duration (range), 21 days (1–112)] resulted in high clinical success in patients with S. aureus (88.4%; MRSA, 80.0%) and CoNS (81.1%) infections, with an overall success rate of 83.3%. Clinical success rate was high (90.0%) in patients who received daptomycin dose >6 mg/kg/day. Overall clinical success rate in patients followed for up to 2 years was 90.7%. AEs and serious AEs possibly related to daptomycin were reported in 6 (3.0%) and 4 (2.0%) patients, respectively. Conclusions: Daptomycin treatment was effective and well tolerated with a sustained response in patients with IE who underwent heart valve surgery. A trend towards better clinical outcomes was observed with higher daptomycin doses.

scholarly journals Home-Based Therapy After Stroke Using the Hand Spring Operated Movement Enhancer (HandSOME II)

2021 ◽  
Vol 15 ◽  
Author(s):  
Rafael Casas ◽  
Melissa Sandison ◽  
Diane Nichols ◽  
Kaelin Martin ◽  
Khue Phan ◽  
...  
Keyword(s):  
Functional Connectivity ◽  
Real World ◽  
Brain Plasticity ◽  
Low Cost ◽  
Biomechanical Testing ◽  
Chronic Stroke ◽  
Reaching Task ◽  
Home Based ◽  
Before And After ◽  
Task Practice

We have developed a passive and lightweight wearable hand exoskeleton (HandSOME II) that improves range of motion and functional task practice in laboratory testing. For this longitudinal study, we recruited 15 individuals with chronic stroke and asked them to use the device at home for 1.5 h per weekday for 8 weeks. Subjects visited the clinic once per week to report progress and troubleshoot problems. Subjects were then given the HandSOME II for the next 3 months, and asked to continue to use it, but without any scheduled contact with the project team. Clinical evaluations and biomechanical testing was performed before and after the 8 week intervention and at the 3 month followup. EEG measures were taken before and after the 8 weeks of training to examine any recovery associated brain reorganization. Ten subjects completed the study. After 8 weeks of training, functional ability (Action Research Arm Test), flexor tone (Modified Ashworth Test), and real world use of the impaired limb (Motor Activity Log) improved significantly (p < 0.05). Gains in real world use were retained at the 3-month followup (p = 0.005). At both post-training and followup time points, biomechanical testing found significant gains in finger ROM and hand displacement in a reaching task (p < 0.05). Baseline functional connectivity correlated with gains in motor function, while changes in EEG functional connectivity paralleled changes in motor recovery. HandSOME II is a low-cost, home-based intervention that elicits brain plasticity and can improve functional motor outcomes in the chronic stroke population.

scholarly journals Description and clinical validation of a real-time AI diagnostic companion for fetal ultrasound examination

2021 ◽  
Author(s):  
Julien Stirnemann ◽  
Remi Besson ◽  
Emmanuel Spaggiari ◽  
Sandra Rojo ◽  
Frederic Loge ◽  
...  
Keyword(s):  
Success Rate ◽  
Real Time ◽  
Ultrasound Examination ◽  
Correct Diagnosis ◽  
Real Life ◽  
Real Data ◽  
Fetal Ultrasound ◽  
Postmortem Diagnosis ◽  
Knowledge Based ◽  
Phenotype Model

Objective: To describe a real-time decision support system (DSS), named SONIO, to assist ultrasound-based prenatal diagnosis and to assess its performance using a clinical database of precisely phenotyped postmortem examinations. Population and Methods: This DSS is knowledge-based and comprises a dedicated thesaurus of 294 syndromes and diseases. It operates by suggesting, at each step of the ultrasound examination, the best next symptom to check for in order to optimize the diagnostic pathway to the smallest number of possible diagnoses. This assistant was tested on a single-center database of 251 cases of postmortem phenotypes with a definite diagnosis. Adjudication of discordant diagnoses was made by a panel of external experts. The primary outcome was a target concordance rate >90% between the postmortem diagnosis and the top-7 diagnoses given by SONIO when providing the full phenotype as input. Secondary outcomes included concordance for the top-5 and top-3 diagnoses; We also assessed a '1-by-1' model, providing only the anomalies sequentially prompted by the system, mimicking the use of the software in a real-life clinical setting. Results: The validation database covered 96 of the 294 (32.65%) syndromes and 79% of their overall prevalence in the SONIO thesaurus. The adjudicators discarded 42/251 cases as they were not amenable to ultrasound based diagnosis. SONIO failed to make the diagnosis on 7/209 cases. On average, each case displayed 6 anomalies, 3 of which were considered atypical for the condition. Using the 'full-phenotype' model, the success rate of the top-7 output of Sonio was 96.7% (202/209). This was 91.9% and 87.1% for the top-5 and top-3 outputs respectively. Using the '1-by-1' model, the correct diagnosis was within the top-7, top-5 and top-3 of SONIO's output in 72.4%, 69.3% and 63.1%. Conclusion: Sonio is a robust DSS with a success-rate >95% for top-7 ranking diagnoses when the full phenotype is provided, using a large database of noisy real data. The success rate over 70% using the '1-by-1' model was understandably lower, given that SONIO's sequential queries may not systematically cover the full phenotype.

scholarly journals Grounding of Human Environments and Activities for Autonomous Robots

2017 ◽  
Author(s):  
Muhannad Alomari ◽  
Paul Duckworth ◽  
Nils Bore ◽  
Majd Hawasly ◽  
David C. Hogg ◽  
...  
Keyword(s):  
Natural Language ◽  
Real World ◽  
Human Activities ◽  
Autonomous Robots ◽  
Proof Of Concept ◽  
Learning Framework ◽  
Art Object ◽  
Simple Sentences ◽  
Robotic Applications ◽  
Continual Learning

With the recent proliferation of human-oriented robotic applications in domestic and industrial scenarios, it is vital for robots to continually learn about their environments and about the humans they share their environments with. In this paper, we present a novel, online, incremental framework for unsupervised symbol grounding in real-world, human environments for autonomous robots. We demonstrate the flexibility of the framework by learning about colours, people names, usable objects and simple human activities, integrating state-of-the-art object segmentation, pose estimation, activity analysis along with a number of sensory input encodings into a continual learning framework. Natural language is grounded to the learned concepts, enabling the robot to communicate in a human-understandable way. We show, using a challenging real-world dataset of human activities as perceived by a mobile robot, that our framework is able to extract useful concepts, ground natural language descriptions to them, and, as a proof-of-concept, generate simple sentences from templates to describe people and the activities they are engaged in.

Sign in / Sign up

Export Citation Format

Share Document