Research on Automated Plankton Identification

When researching automatic methods for identification its important to set a benchmarks for human performance. I've been studying this topic for some time, and realised that I should share this paper - Culverhouse et al. (2014) An empirical assessment of the consistency of taxonomic identifications, Marine Biology Research, 10:1, 73-84, DOI: 10.1080/17451000.2013.810762.

It reports a study of 21 skilled plankton analysts drawn from 6 international laboratories. Its main conclusion is that humans can be variable in performance, with only 14 of the 21 experts, in a series of experiments, returning self-consistency scores better than 90%, the remaining 7 were ranging from almost random decision makers to no better than 73% self-consistent. The paper is available in the SCOR WG-130 dropbox available here

Publications Shared with Dropbox

Thought I'd posted a call for PhD students interested in machine identification of natural objects. I'll try again! I am offering a PhD topic under the call described below. Its on machine identification of benthic flora and fauna. I am seeking a person with good maths abilities, perhaps a knowledge of machine vision, perhaps using OpenCV. The work will involve a new feature detector and texture analysis, and machine learning. We have excellent resources for this project including access to a large data set of plankton, fish and benthic organisms. The work will eventually be embedded in an AUV that we have at Plymouth. Email me for further details.
[email protected].

PhD Studentships – Centre for Robotics and Neural Systems (CRNS), Plymouth University, UK

Following on the exceptional performance of Plymouth University in the UK Research Excellence Framework 2014, the university is now advertising the availability of 50 PhD studentships for study commencing in 2015/16.
Of these studentships, at least two will be allocated to research projects and supervisors of the Centre for Robotics and Neural Systems (CRNS), within the Faculty of Science and Engineering.
The first studentship can cover any topic within the CRNS research areas:
Cognitive and Humanoid Robotics
Computational Neuroscience
Artificial Intelligence
The second studentship will focus on Autonomous Systems, possibly with relevance to unmanned marine systems, and will be in collaboration with the School of Marine Science and Engineering:

For more details on the CRNS research projects, interests and supervisors, see:
https://www1.plymouth.ac.uk/research/crns/Pages/default.aspx

For general information on the studentships and application procedure, see:

https://www.plymouth.ac.uk/student-life/your-studies/the-graduate-school/fully-funded-phd-studentships

We recommend that in the application form and cover letter you specify a research topic and preferred supervisor. You are also encouraged to contact the individual supervisors before applying, for informal discussion.

The application deadline is 15 September 2015

Centre for Robotics and Neural Systems Angelo Cangelosi has been awarded a contract from the US Air Force Office of Science and Research (AFSOR-EOARD) to work on trust in human-robot interaction. The $540k contract will see the hiring of two 4-year PhD students, one designing the social skills of the robots and the otehr carrying out exp…

Karenia brevis samples analyzed by FlowCAM at the 6th Symposium on Harmful Algae taking place in Austin, Texas Nov 13-17. This harmful algae is present along the Texas coast in a prolonged "Red Tide".