Dr Carl Chalmers

School of Computer Science and Mathematics

Poaching the Poacher: The development of Artificial Intelligence and practical skills in the curriculum as well job opportunities in conservation and Artificial Intelligence.

There is a growing need for graduates who are equipped with practical skills in the area of Artificial Intelligence (AI) and Data Science. These skills are high on international government agendas and there is a strategic push to compete with world leaders in AI services and innovation. The field is both diverse and broad but at LJMU we have particular expertise in applied computer vision and Deep Learning (DL) which has made us one of the world leaders in its application to global conservation.

Academics have developed a world class Conservation AI platform that is foundational in conservation technology. It is being used worldwide in South Africa, Malaysia, Indonesia, Tanzania and Sudan by the BBC, WWF, Knowsley Safari Park, Chester Zoo, Endangered Wildlife Trust, AUDUBON CANYON RANCH, Greater Mahale Ecosystem Research and Conservation Team and the Northern Jaguar Project. Many of the challenges highlighted by these organisations are embedded within our research informed teaching that is delivered at MSc and PhD level. Coursework’s and projects are designed to tackle the challenges raised by our partners. For example, several of problems solved by the academic team (e.g., animal classification and behavioural analytics) are used to challenge students in their coursework and projects. This an effective way of providing students with the necessary skills that are required to solve their own problems in industry or advanced research.

The work carried out by the academic team has been reported BBC Radio, ITV Granada Reports, Nvidia GTC and New Scientist. To date conservationists have used the platform to process over 460,000 images and identify over 90,000 animals. This has aided conservationists in their studies and dissemination activities. One of our more recent successes is a preferred supplier of ConservationAI in studies commissioned by Animal and Plant Health Agency (APHA) which is part of DEFRA.

Impact curriculum:

The applied research carried out by Dr Carl Chalmers and Prof Paul Fergus has resulted in a new validated MSc Artificial Intelligence (Machine Learning) programme which started September 2020). This programme builds on the academics industrial and research experience and provides material that allows student to develop practical enterprise level skills. This course was developed with input from Nvidia and other industrial experts and has proven to be essential for graduate employability in the field of AI.

The lecture and lab material were heavily underpinned with the academic team’s expertise in industrial software development and applied AI research. This allows students to develop end-to-end AI solutions that can deployed and used to solve real-world problems. Students complete the course with the ability to effectively communicate complex technical ideas and deliver robust technical solutions for a number of given problems. Coursework’s are split across written assessments in the form of academic papers and practical development tasks using cutting edge approaches.

Alongside the development of the MSc programme the academics wrote a supporting 450-page book entitled “Applied Deep Learning: Tools, Techniques and Implementation (ISBN:3031044193/9783031044199). In the final MSc project students are paired with external partners to work on real-world problems through which the academic team have strong research partners.

Broader Change:

Successful students have gone on to successfully secure employment with large blue-chip organisations which include IBM and Microsoft. The academic team have received the following emails from pervious students:

“I just wanted to say thanks for all of yours and Pauls help in my masters, as it’s now meant I’ve been able to excel in my current job”.

“Thanks again for everything you and Paul have done for, I’m very privileged to be in the situation I am currently, and I think that one of the reasons for my boss’s interest in the course is because I have spoken so highly of the quality of it”.

“Hope you are both well. Just thought I would let you both know that from leaving in September I have started a new job as a data analyst. However, I have now also been offered two different graduate schemes, at CGI in Manchester and The Office for Nuclear Regulation in Liverpool (Bootle). I now have a decision to make as to which one I decide to join, but I would like to thank you both for your expertise over the past year. The content we covered really helped me throughout the assessment centers and interviews. Even though It was more difficult for all of us to complete the course over zoom because of the restrictions of covid, I believe I would not have these opportunities without it. So thanks again”.

Some examples of the work is shown below:

A collection of images taken at night of animals and other shapes such as trees.  The image includes a graphical overlay which shows that the AI has correctly identified which of the shapes are animals.