PhD Oral Exam - Elahe Rahimian Najafabadi, Information and Systems Engineering
Myoelectric Control for Active Prostheses via Deep Neural Networks and Domain Adaptation
This event is free
School of Graduate Studies
Recent advances in Biological Signal Processing (BSP) and Machine Learning (ML), in particular, Deep Neural Networks (DNNs),
When studying for a doctoral degree (PhD), candidates submit a thesis that provides a critical review of the current state of knowledge of the thesis subject as well as the student’s own contributions to the subject. The distinguishing criterion of doctoral graduate research is a significant and original contribution to knowledge.
Once accepted, the candidate presents the thesis orally. This oral exam is open to the public.
Abstracthave paved the way for development of advanced Human-Machine Interface (HMI) systems for decoding human intent and controlling artificial limbs. Myoelectric control, as a subcategory of HMI systems, deals with detecting, extracting, processing, and ultimately learning from Electromyogram (EMG) signals to command external devices, such as hand prostheses. In this context, hand gesture recognition/classification via Surface Electromyography (sEMG) signals has attracted a great deal of interest from many researchers. This is mainly due to its high potential for improving the quality of control over the actions of prostheses, which can significantly enhance the quality of lives of hand amputated individuals.
Despite extensive progress in the field of myoelectric prosthesis, however, there are still limitations that should be addressed to achieve a more intuitive upper limb prosthesis. Through this Ph.D. thesis, first, we perform a literature review on recent research works on pattern classification approaches for myoelectric control prosthesis to identify challenges and potential opportunities for improvement. Then, we aim to enhance the accuracy of myoelectric systems, which can be used for realizing an accurate and efficient HMI for myocontrol of neurorobotic systems. In this regard, we design and implement two novel DNN models to improve the overall performance and accuracy of the Hand Gesture Recognition (HGR) system. Beside improving the accuracy, decreasing the number of parameters in DNNs plays an important role in a HGR system. More specifically, a key factor to achieve a more intuitive upper limb prosthesis is the feasibility of embedding DNN-based models into prostheses controllers. We propose to solve this problem by designing a DNN-based architecture that utilizes depthwise separable convolutions and adaptive average pooling, resulting in a less complex network. Moreover, capitalizing on the recent advancements of the attention-based architecture, we develop an attention-based framework for processing sEMG signals. On the other hand, transformers are considered to be powerful DNN models that have revolutionized the Natural Language Processing (NLP) field and showed great potentials to dramatically improve different computer vision tasks. Therefore, we propose a Transformer-based neural network architecture to classify and recognize upper-limb hand gestures. Finally, another goal of this thesis is to design a modern DNN-based gesture detection model that relies on minimal training data while providing high accuracy. Although DNNs have shown superior accuracy compared to conventional methods when large amounts of data are available for training, their performance substantially decreases when data are limited. Collecting large datasets for training may be feasible in research laboratories, but it is not a practical approach for real-life applications. We propose to solve this problem, by designing a framework which utilizes a combination of temporal convolutions and attention mechanisms. We provide a novel venue for adopting few-shot learning, to not only reduce the training data, but also to eventually mitigate the significant challenge of variability in the characteristics of sEMG signals. In other words, the proposed framework allows a myoelectric controller, that has been built based on background data, to adapt to the changes in the stochastic characteristics of sEMG signals using a small number of new observations.