Computational guidance to help tackle human-wildlife conflict
Dr Kartic Subr will be working alongside conservationists on a project aimed at modelling spatio-temporal populations of certain species of animals.
Dr Subr won a £100,000 grant from the Royal Society for his CAT-SPATS project, which addresses the escalating problem of human-wildlife conflict, focusing on trends in India.
The central hypothesis in this project is that the effectiveness of conservation efforts in India may be significantly increased with the aid of computational guidance to experts in the field. The goal of this project is to estimate the spatio-temporal distributions of certain species, such as leopard, using pictures of animals captured by automatic camera traps. The expected outcome is automatic classification of images from camera-traps, which will then be used to estimate spatio-temporal distributions. Much of this data is currently gathered using ad hoc schemes and manually processed. The availability of computational schemes will tackle the problem of scale and enable conservationists in the field to warn local populations of imminent threats, determine anti-conflict schemes, target anti-poaching efforts, drive policies on issues such as where to build highways, bridges, etc.
Thousands of people and hundreds of wild animals are killed every year in India as dwindling natural wild habitats become increasingly fragmented, blending with rural and urban sprawls. The struggle for survival on both sides of the boundaries of forests in India is a source of increasing tension, plaguing the efforts of conservationists who toil hard to restore harmony. Policy makers in developing countries are torn between urgent decisions necessary for the immediate upliftment of the poor and important decisions such as reducing environmental impact. In countries such as India, this indecision is exacerbated by the unavailability of temporally-accurate data due to the enormous rate of change. Rapid, yet accurate estimation of statistics, are vital for informed decisions to be made.
For example, Bangalore (the capital of Karnataka state) has struggled to cope with a 60% increase in population (5million to 8m) over the past decade. The abrupt expansion of Bangalore city has had an adverse effect on the surrounding wilderness. There is escalating tension between man and wildlife on the boundaries of such settlements  leading to a loss of human lives, mass destruction of property and crops and the deaths of threatened species. The undercurrent of competition between man and animal destabilises urban and rural areas in the region.
Dr Subr says:
“I am excited by this first opportunity to establish connections with conservationists. I expect that the collaboration will serve as a platform to gather data across a wider range of problems in the future.”
Dr Kartic Subr is a Senior Lecturer in Computer Graphics at the School of Informatics, University of Edinburgh.