0% Complete
صفحه اصلی
/
سی و سومین کنفرانس بین المللی مهندسی برق
Optimizing Dual IMU Sensor Placement for Gait Phase Detection with LSTM Models
نویسندگان :
Mahya Abedi
1
Zolfa Anvari
2
Hamed Ghafarirad
3
Mohammad Zareinejad
4
1- دانشگاه صنعتی امیرکبیر
2- دانشگاه صنعتی امیرکبیر
3- دانشگاه صنعتی امی
4- دانشگاه صنعتی امیر
کلمات کلیدی :
Gait phase detection،IMU sensors،sensor placement evaluation،LSTM network،real-time optimization،quantization techniques
چکیده :
Precise analysis of human gait phases is essential for diagnosing motor disorders, optimizing athletic performance, and designing assistive devices. However, achieving high accuracy in gait phase detection while maintaining real-time applicability poses significant challenges, primarily due to suboptimal sensor placement and latency constraints. This study addresses these challenges by optimizing the placement of Inertial Measurement Unit (IMU) sensors and applying quantization techniques to enhance real-time performance. Using a Long Short-Term Memory (LSTM) algorithm, four key gait phases—heel strike, toe strike, heel off, and toe-off—were analyzed across five distinct sensor configurations. Training features were uniquely selected for each configuration based on a systematic feature selection process to ensure optimal model performance. Experiments under controlled walking conditions demonstrated that the placement of sensors on the shin and bridge of the foot achieved the highest accuracy (97.4%), followed by the thigh and bridge of the foot (96.34%). Additionally, quantization techniques reduced latency to 3–5 milliseconds without compromising accuracy, highlighting the feasibility of real-time applications. The findings of this study establish a foundation for selecting optimal sensor placements tailored to specific accuracy and performance requirements. Researchers and practitioners can leverage these configurations to compare detection accuracy against their unique application needs, facilitating the design of efficient and practical gait analysis systems for medical, rehabilitation, and sports contexts.
لیست مقالات
لیست مقالات بایگانی شده
Area-Efficient Partially-Pipelined Architecture for Fast-SSC Decoding of Polar Codes
Mehdi Saeidi - Matin Hashemi
Solving the inverse problem for EEG signals when learning a new motor task using GRU neural network
Milad Khosravi - Fariba Bahrami - Behzad Moshiri - Ahmad Kalhor
پیشبینی بازار سرمایه به کمک دادهکاوی با الگوریتمهای رگرسیونی
شیوا نمایان - محمدشهرام معین
An Improved Version of the SIPO Algorithm with Fast Convergence Speed
Amir Soltany Mahboob - Hadi Shahriar Shahhoseini - Mohammad Reza Ostadi Moghaddam - Shima Yousefi
High Step up DC/DC Converter with Low Input Current Ripple and Low Voltage Stress on Semiconductors
Saed Mahmoud Alilou - Mohammad Maalandish - Soheil Nouri - Seyed Hossein Hosseini
Effect of Physical Characteristics on Artificial Neural Network Error Reduction for Indoor Propagation Modeling
SeyedehMounes Eslami - Amir Ahmad Shishegar
Fault tolerant control design for linear systems based on cubic observers
Mahsa Hasanshahi - Malihe Maghfoori Farsangi - Elham Amini Boroujeni
A Siamese Neural Network for Predicting snoRNA-Disease Association
Milad Besharatifard - Fatemeh Zare-Mirakabad
Secrecy Sum Rate Analysis and Power Allocation with OSTBC and Artificial Noise for MIMO Systems
Abdolrasoul Sakhaei Gharagezlou - Mahdi Nangir - Nima Imani - Amir Poorfaraj Liqvan
True Class-E Design For Inductive Coupling Wireless Power Transfer Applications
Ali Asghar Razavi Haeri - Aminghasem Safarian - Ali Fotowat-Ahmady
بیشتر
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 41.7.4