Personal tools

Difference between revisions of "ML based Quantitative Movement Analysis on a Portable IoT Camera (1-2S/B)"

From iis-projects

Jump to: navigation, search
 
(7 intermediate revisions by the same user not shown)
Line 5: Line 5:
  
 
==Short Description==
 
==Short Description==
This project focusses on the development of an unobtrusive multisensory embedded system to assist coaches to better quantify jumping trajectories of athletes. Within the short duration of a ski-jump (< 10 seconds) and exposed to the conditions of nature (snow, wind, temperature) athletes must solve extremely difficult optimisation problems. Flight trajectories of athletes are decisive for victory in a ski jumping competition. They are influenced by the properties of the inrun, the take-off speeds, the applied forces, the athletes’ body position as well as ski edging angles during flight.  
+
Muscles and tendons within the muscle-tendon unit have different properties and respond differently to external loads and training stimuli. Imbalances in muscle strength or tendon stiffness, for example, can affect movement efficiency and lead to injuries. In clinical biomechanics, understanding changes in muscle and tendon properties is vital for creating effective treatment plans.
  
The challenge in this project lies in the combination and synchronization of the sensors and the wireless data transmission between the flying athlete and the coaching tower. In addition, due to the complexity of such a flight situation, the body-mounted sensors and devices must be tiny and barely perceptible to the athlete so as not to disturb his/her sensitive jumping system.  
+
However, studying muscle and tendon mechanics during movement often requires multiple sensors such as 3D motion capture, ultrasound, electromyography, and force measurements. Obtaining this data can be challenging, particularly during fast movements like running, due to high acceleration and perspiration (falling off sensors), or when working with specific patient groups, such as children with cerebral palsy.
 +
 
 +
The goal of this work is to adapt a trained machine learning model that can estimate body positions and calculate changes in the muscle-tendon unit lengths to work on a portable, smart IoT camera. This could impact the way biomechanical measurements are performed today, as the proposed system should be easy and quick to use (e.g., time is key for testing patients in clinics) and inexpensive compared to traditional necessary hardware.
  
 
===Status: Available ===
 
===Status: Available ===
: This project will be in cooperation with the Polish National Ski Jumping team.
+
: Students will be co-supervised by the Center for Project Based Learning.
: Students will be co-supervised by the Center of Project Based Learning.
 
 
: Looking for 1-2 Semester/Master students
 
: Looking for 1-2 Semester/Master students
: Contact: [[:User:Cleitne | Christoph Leitner]], [mailto:schuluka@student.ethz.ch Lukas Schulthess (PBL)]
+
: Contact: [[:User:Cleitne | Christoph Leitner]], [mailto:marco.giordano@pbl.ee.ethz.ch Marco Giordano (PBL)]
  
 
===Prerequisites===
 
===Prerequisites===
: Embedded systems and PCB design
+
: Python
: Microcontrollers
+
: Machine Learning on Microcontrollers
 
<!--  
 
<!--  
 
===Status: Completed ===
 
===Status: Completed ===
Line 26: Line 27:
 
===Status: In Progress ===
 
===Status: In Progress ===
 
: Student A, StudentB
 
: Student A, StudentB
: Supervision: [[:User:Cleitne | Christoph Leitner]], Lukas Schulthess (PBL- ETHZ)
+
: Supervision: [[:User:Cleitne | Christoph Leitner]], [mailto:marco.giordano@pbl.ee.ethz.ch Marco Giordano (PBL)]
 
--->
 
--->
  
 
===Character===
 
===Character===
 
: 10% Literature research
 
: 10% Literature research
: 20% Sensor interfaces
+
: 60% tinyML
: 35% Embedded System Design
+
: 30% Interface engineering
: 35% Wireless Communication
 
  
 
===Professor===
 
===Professor===
Line 46: Line 46:
  
 
==Detailed Task Description==
 
==Detailed Task Description==
The main goal of this thesis is to design, build and test a multisensory system to determine jump trajectories of athletes. The sensors should be able to record ski edging angles (e.g., using IMUs) and insole pressures (e.g. using piezoresistive sensors) during approach, take-off, and landing. The acquired data should be sent from the senor node to the coaching tower. According to the level of the student and the chosen thesis type (MT/BT/ST) the work will include some or all following tasks:
+
This research aims to transfer an existing machine learning model and an accompanying post-processing algorithm to an ML accelerator, e.g., a Google Coral device. The accelerator will be paired with a portable IoT camera, and the algorithm's predictions will be sent via, e.g., an USB cable to a sub-computer. The model will be tested under lab conditions using data from walking or running subjects.
[[File:SkiEdgingAngles.png|thumb|]]
 
  
 
===Goals===
 
===Goals===
=====Sensors and Acquisition=====
+
* Research of available ML accelerators and cameras
* Investigation and evaluation of various commercially available sensor technologies (IMUs, pressure sensors).
+
* Familiarization with ML accelerator
* Evaluation of the attachment of sensors to skis and in boots.
+
* Model quantization from tensorflow to tensorflow lite
* Use an in-house multipurpose embedded systems controller (Vitalcore) to build & test data collection with sensors.
+
* Deployment of the model on the accelerator and testing
=====Communication and Data Transfer=====
+
* Build interfaces between IoT camera (2d video capture), the accelerator (prediction) and the sub-computer (display)
* Develop a data transfer strategy to  
+
* Testing of the implemented pipeline on real data
** collect data from two different sensors, and  
 
** to transmit synchronized (raw) data to the coaching tower.
 
** Use an in-house multipurpose embedded systems controller (Vitalcore) to test data transfers via BLE.
 
** Test data transfer in ski jumping arena situation and revaluate transfer strategy if necessary.
 
=====Assembly and Test=====
 
* Make a PCB board design for the readout system.
 
* Design a casing to attach sensors system on the skis or shoes. Aiming for a minimalistic form-factor and weight.
 
* Test, build and evaluate a working prototype in laboratory conditions and in a real-life environment.
 
 
 
  
 
===Practical Details===
 
===Practical Details===
Line 84: Line 74:
 
[[Category:Wearables for Sports and Fitness]]
 
[[Category:Wearables for Sports and Fitness]]
 
[[Category:SmartSensors]]
 
[[Category:SmartSensors]]
 +
[[Category:WearablesSF_signals]]
 
[[Category:Available]]
 
[[Category:Available]]
 
[[Category:Group Work]]
 
[[Category:Group Work]]
 
[[Category:Semester Thesis]]
 
[[Category:Semester Thesis]]
 
[[Category:Master Thesis]]
 
[[Category:Master Thesis]]
 +
[[Category:Cleitne]]
 +
  
  

Latest revision as of 11:12, 23 July 2023

Running.gif



Short Description

Muscles and tendons within the muscle-tendon unit have different properties and respond differently to external loads and training stimuli. Imbalances in muscle strength or tendon stiffness, for example, can affect movement efficiency and lead to injuries. In clinical biomechanics, understanding changes in muscle and tendon properties is vital for creating effective treatment plans.

However, studying muscle and tendon mechanics during movement often requires multiple sensors such as 3D motion capture, ultrasound, electromyography, and force measurements. Obtaining this data can be challenging, particularly during fast movements like running, due to high acceleration and perspiration (falling off sensors), or when working with specific patient groups, such as children with cerebral palsy.

The goal of this work is to adapt a trained machine learning model that can estimate body positions and calculate changes in the muscle-tendon unit lengths to work on a portable, smart IoT camera. This could impact the way biomechanical measurements are performed today, as the proposed system should be easy and quick to use (e.g., time is key for testing patients in clinics) and inexpensive compared to traditional necessary hardware.

Status: Available

Students will be co-supervised by the Center for Project Based Learning.
Looking for 1-2 Semester/Master students
Contact: Christoph Leitner, Marco Giordano (PBL)

Prerequisites

Python
Machine Learning on Microcontrollers

Character

10% Literature research
60% tinyML
30% Interface engineering

Professor

Luca Benini

↑ top

Detailed Task Description

This research aims to transfer an existing machine learning model and an accompanying post-processing algorithm to an ML accelerator, e.g., a Google Coral device. The accelerator will be paired with a portable IoT camera, and the algorithm's predictions will be sent via, e.g., an USB cable to a sub-computer. The model will be tested under lab conditions using data from walking or running subjects.

Goals

  • Research of available ML accelerators and cameras
  • Familiarization with ML accelerator
  • Model quantization from tensorflow to tensorflow lite
  • Deployment of the model on the accelerator and testing
  • Build interfaces between IoT camera (2d video capture), the accelerator (prediction) and the sub-computer (display)
  • Testing of the implemented pipeline on real data

Practical Details


Links

↑ top