Recherche partenaire principal
Human Activity Recognition is already a mature area and can now be extended to the animal kingdom. This project will use video cameras to observe animal behaviour, either as pets, as domestic animals or in the wild. Extending human activity recognition for interpreting animal behaviour is not straightforward at all. Video sequences would be processed and analysed using machine learning then animal behaviour would be interpreted and classified. Anomalous behaviour will be identified as a result. This project will have an impact on analysing, understanding and interpreting behaviour of pets, domesticated animals as well as animals in the wild. Our expertise lies in video processing, machine learning and human activity recognition from video sequences. I am currently the leader of the project "Human Activity Recognition from Video Sequences," funded by Dstl, UK.