Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userDeveloping an Intelligent Controller for lower limb Neuroprosthesis Part I:Designing of The Training Data
تطوير محاكي بيانات التدريب لمتحكم معوض عصبي للأطراف السفلية
1535
1 35
0
(
0
)
Created by
Shamra Editor
Ask ChatGPT about the research
تصمم المعوضات العصبية لاستعادة الوظائف الحركية المفقودة لدى مرضى شلل الأطراف السفلية بالاعتماد على الإثارة الكهربائية الوظيفية. إِذ تحدد متحكمات المعوضات العصبية العلاقة بين نبضات الإثارة المطبقة و زوايا المفاصل من أجل توليد أنماط الإثارة المناسبة للحركات المطلوبة. و يحتاج تطوير المتحكمات الذكية إلى بناء نموذج عكسي باستخدام مجموعة بيانات التدريب المؤلفة من دخل (نبضات الإثارة) و خرج (زوايا المفاصل) التي يتم الحصول عليها بأساليب تجريبية. بسبب كثرة التجارب المزعجة للمريض و الحاجة لتكرارها خلال مراحل تصميم المتحكمات، يستخدم هذا البحث النمذجة و المحاكاة لتوليد مجموعة بيانات دخل (نبضات الإثارة)/خرج (زوايا المفاصل) من خلال: تطوير نموذج لجسم الإنسان، و محاكاة تجارب تطبيق الإثارة الكهربائية لمجموعة عضلات مربعة الرؤوس الفخذية الباسطة للركبة خلال تأرجح الساق. ربطت الدراسة بين ثلاثة برامج لتطوير نموذج جسم الإنسان: برنامج Visual Nastran 4D لبناء نموذج أجزاء الجسم وبرنامج Virtual Muscle 4.0.1 لنمذجة العضلات الباسطة للركبة، و مكتبة Simulink لتمثيل خصائص المفصل غير الفعالة، ثم استخدم النموذج في محاكاة تعيين مجموعة البيانات من خلال تطبيق إشارات جيبية و عشوائية لإثارة العضلات الباسطة للركبة. يلائم النموذج المطور مختلف مستخدمي المعوضات العصبية، بسبب استخدامه لبرامج ذات واجهات مستخدم رسومية تسمح بتعديل بارامترات جسم الإنسان و العضلات بسهولة، مما يجعله نموذجاً عاماً، يستخدم في تطوير المتحكمات لاستعادة الحركات المفقودة مثل النهوض، و المشي، و القفز و غيرها....و لأن نتائج المحاكاة تشابه نتائج التجارب العملية، يقلل استخدام النموذج المطور عدد التجارب المنفذة على المرضى خلال عملية تصميم متحكمات المعوضات العصبية.
Neuroprosthesis can be used to restore lost motor functions for paraplegicsby using functional electrical stimulation (FES). Neuroprosthesis controllers determine the relationship between the stimulation pulses and joint angles to generate electrical stimulation patterns for the desired movement.To develop intelligent controllers, an inverse model which is the basic component of the intelligent controller is built by using empirical approaches to get a data set that consists of input (stimulation pulses) and output (joint angles). Because of the numerous exhausting experiments on patients and the need for repetition during Controller design, this study uses modeling and simulation to generate the data setthrough developing humanoid model, and simulating practical trials of quadriceps stimulation during swing leg movement. We connected three programs to develop a humanoid model by building: body segments in Visual Nastran 4D, muscles in Virtual Muscle 4.0.1, and passive joint properties in Matlab/ Simulink. Then the humanoid model was used to producethe identification data sets, through applying sinusoidal and random signals to simulate the stimulation of the knee extensors. The humanoid model can fit different users by using a number of graphical user interface screens to change the human and muscles parameters, so it is a generic model. It can be used in developing controllers to restore lost movement such as standing up, walking, jumping, etc. The simulation results is similar to practical trials, so using the developed model can reduce the number of experimental tests to be performed with patients during Neuroprosthesis controllers design.
References used
Braz, G. P., Russold, M., & Davis, G. (2009). "Functional Electrical Stimulation Control of Standing and Stepping After Spinal Cord Injury: A Review of Technical Characteristics". In Neuromodulation: Technology at the Neural Interface,International Neuromodulation Societyvol. 12, pp. 180-190
Horch, K., & Dhillon, G. (2004). "NEUROPROSTHETICS: Theory and Practice". World Scientific Publishing Co
Afzal, T., Khan, L., & Tokhi, M. (2010). "Simulation of a Patient Driven Strategy for FES Supported Sit-to-stand Movement". International Conference on Information and Emerging Technologies (ICIET). Pakistan: IEEE
Chang, G.-C., Luh, J.-J., Liao, G.-D., Lai, J.-S., & Cheng, C.-K. (1997). "A Neuro-Control System for theKnee Joint Position Control with Quadriceps Stimulation". IEEE Transaction on rehabiltation engineering, vol. 5, pp. 2- 11
Ferrarin, M., Palazzo, F., Riener, R., & Quintern, J. (2001). "Model-Based Control of FES-Induced Single Joint Movements". IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 9, pp. 245 -257
rate research
Read More
Machine translation of user-generated code-mixed inputs to English is of crucial importance in applications like web search and targeted advertising. We address the scarcity of parallel training data for training such models by designing a strategy o
f converting existing non-code-mixed parallel data sources to code-mixed parallel data. We present an m-BERT based procedure whose core learnable component is a ternary sequence labeling model, that can be trained with a limited code-mixed corpus alone. We show a 5.8 point increase in BLEU on heavily code-mixed sentences by training a translation model using our data augmentation strategy on an Hindi-English code-mixed translation task.
Unsupervised neural machine translation (UNMT) that relies solely on massive monolingual corpora has achieved remarkable results in several translation tasks. However, in real-world scenarios, massive monolingual corpora do not exist for some extreme
ly low-resource languages such as Estonian, and UNMT systems usually perform poorly when there is not adequate training corpus for one language. In this paper, we first define and analyze the unbalanced training data scenario for UNMT. Based on this scenario, we propose UNMT self-training mechanisms to train a robust UNMT system and improve its performance in this case. Experimental results on several language pairs show that the proposed methods substantially outperform conventional UNMT systems.
In this paper, it has
merged two techniques of the artificial intelligent, they are the
ants colony optimization algorithm and the genetic algorithm, to
The recurrent reinforcement learning trading system
optimization. The proposed trading system
is based on an ant
colony optimization algorithm and the genetic algorithm to
select an optimal group of technical indicators, and fundamental
indicators.
Extracting keyphrases that summarize the main points of a document is a fundamental task in natural language processing. Supervised approaches to keyphrase extraction(KPE) are largely developed based on the assumption that the training data is fully
annotated. However, due to the difficulty of keyphrase annotating, KPE models severely suffer from incomplete annotated problem in many scenarios. To this end, we propose a more robust training method that learns to mitigate the misguidance brought by unlabeled keyphrases. We introduce negative sampling to adjust training loss, and conduct experiments under different scenarios. Empirical studies on synthetic datasets and open domain dataset show that our model is robust to incomplete annotated problem and surpasses prior baselines. Extensive experiments on five scientific domain datasets of different scales demonstrate that our model is competitive with the state-of-the-art method.
We present new state-of-the-art benchmarks for paraphrase detection on all six languages in the Opusparcus sentential paraphrase corpus: English, Finnish, French, German, Russian, and Swedish. We reach these baselines by fine-tuning BERT. The best re
sults are achieved on smaller and cleaner subsets of the training sets than was observed in previous research. Additionally, we study a translation-based approach that is competitive for the languages with more limited and noisier training data.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
