Daniel Leightley

A blog by Daniel Leightley PhD, a researcher based in the United Kingdom with interests in computer vision, machine learning, big data little data and deep learning.

European Psychiatric Association 18th meeting in Gothenburg

The European Psychiatric Association (EPA) Gothenburg 18th meeting of the Section of Epidemiology and Social Psychiatry aims to to provide a platform for psychiatric epidemiology and social psychiatry research from across the world. I submitted an abstract related to the use of Health Episode Statistics (HES) in epidemiology (more on this later). This year EPA was held in Gothenburg, Sweden between November 30-December 3, 2016 at the Scandic Hotel Opalen, near the centre of Gothenburg. We flew from London Gatwick on Tuesday 29th November in time for registration on Wednesday 30th. I was really excited to attend this conference, it is my first overseas conference with King's College London (KCL). Day 1 - Wed 30 Nov The conference started with...

Postural Stability During Standing Balance and Sit-to-Stand in Master Athlete Runners Compared With Non-Athletic Old and Young Adults

Great news, I have just had my second journal accepted to appear in the Journal of Aging and Physical Activity (JAPA). I am really pleased that the work will soon be published. It forms part of my PhD work, bringing together motion capture (depth sensor) and mobility analysis to evaluate masters athletes. Of particular note is the acceptability by the community to utilise depth sensor technology as the 'gold standard'. This will hopefully spur other researchers to utilise marker-less technology over marker-based solutions. You can find the article here. The aim of the paper study was to compare postural sway during a series of static balancing tasks and during five chair rises between healthy young (mean (SEM) age 26(1)...

Society for Social Medicine: Conference Review

The Society for Social Medicine (SSM) aims to promote the development of scientific knowledge in social medicine. It covers a wide range of topics from epidemiology to health services. I submitted work related to the methodology employed to link electronic healthcare records in England, Scotland and Wales; out of all the work I observed at the conference it was the most technical (more on this later). Each year the Society holds an Annual Scientific Meeting for members and non-members to attend, present work, network and collaborate. This year SSM 2016 was held in York, England, at York University, James College campus from 14th September 2016 to 16th September 2016. The campus is located about 45 minutes walk from York train...

IPDLN 2016: A conference review

IPDLN is an international conference on Population Data Linkage. It takes place every two years. This year IPDLN 2016 was hosted in Swansea, Wales, at the Swansea University Bay campus from 24th August 2016 to 26th August 2016. The campus is located on the sea front, approximately a one hour walk from Swansea train station (I decided to walk it). It has taken me a couple of weeks to post this - been very busy due to a new project starting up. I arrived in Swansea on the 24th August 2016 direct from London. I was really excited because I did not know what to expect from this type of conference. After all, it was my first data linkage conference...

Recording Kinect One Stream using C#

During my PhD I had to obtain three of the data streams (RGB, Depth and Skeleton) of the Kinect One sensor. Sadly, Microsoft, at the time, did not provide any example for recording the raw data itself. They do offer Kinect Studio, however it’s a closed ecosystem without the ability of accessing the raw data. In this blog post I wanted to share with you the C# code snippets for how I recorded the data streams and achieved near 30ftp while saving these streams. You can see an example of the code over on my GitHub page here - please note that the code it is not optimised in any way. The code is very basic and just an...

Face and Eye Detection with Python - Static Image

Detecting the face and eyes of a static image in the frontal plane is a straightforward Computer Vision task. We are able to use the pre-trained cascade models which come with OpenCV; these are based on trained Haar features. I wanted to provide a copy of the syntax I use. Whenever I approach a problem, I always use a basic syntax first, as a baseline comparison. In this post, I will detail how to identify any face and/or eyes using Python and OpenCV using static images. I will make a follow-up blog post for how to identify face and eyes using a web camera in real-time. It is very important that you have both setup on your system, otherwise...

Using MEAN - Choosing the stack for InDEx

Identifying the role that technology plays now and in the future is crucial. We should always seek to future-proof our developments to ensure they remain relevant in the short/medium term. We are at the start of the InDEx project project; where we are mindful of selecting the most suitable technology platform to take us into the future. One of the first decisions we have to make is: which full-stack (otherwise development/production environment) should develop with. For the uninitiated, the term full-stack means working with all the layers from the back-end to the front-end technologies. For many years, the standard environment was referred to as the LAMP and it primarily involved Linux, Apache, MySQL and PHP or Perl or...

InDEx App: A Medical Research Council Funded Project

Information about Drinking for EX-serving personnel (InDEx) InDEx is a project funded by the Medical Research Council to develop an effective electronic alcohol intervention delivered via a hybrid app. The aim is to help people leaving the Armed Forces to lower their alcohol consumption. This is done by sending user-level personalised messages via email, SMS and push-notifications. In addition, the app provides support for the user during the intervention. The InDEx app is being developed at the King's Centre for Military Health Research (KCMHR), and the project is led by Dr Laura Goodwin. The project website is now live at https://www.index-app.org and we hope to be updating the website as the project progresses. Aim and Scope It...

Journal of Biomedical and Health Informatics - Paper Accepted and Published

Great news, my first ever journal paper has been accepted to appear in the Journal of Biomedical and Health Informatics (J-BHI). I am over the moon and very pleased. The paper is the culmination of my PhD, bringing together motion capture, 3D feature representation, big data, little data and machine learning. J-BHI was choosen for submission as its aims and objectives fitted well with our work. The paper focuses on quantification of human motion to support clinicians in the decision-making process. We proposed a framework that automatically recognises and evaluates human mobility impairments using a depth sensor. The framework is composed of two parts. Firstly, it recognises motions, such as sit-to-stand or walking 4 metres, using feature representation techniques and...

Its a new layout and a new blog

I have had this website up and running for more than 5 years. In that time more than 50,000+ unique visitors have landed on the site (according to Google Analytics). However, it was time for a change, time to be more transparent, discuss my work and research. You are now reading this blog on a new platform, with weekly (time permitting) content being published detailing my research, presenting code and general research ramblings. My first blog post accompanying this website will explain why I made the choice to use Ghost, the design and other technical aspects (like https and CloudFlare). I'd love any feedback you may have on any aspect of this post, the site and its general usability....