• Publications
  • Influence
Fully Convolutional Crowd Counting on Highly Congested Scenes
The state-of-the-art for crowd counting in high density scenes is advanced by further exploring the idea of a fully convolutional crowd counting model introduced by (Zhang et al., 2016), and a training set augmentation scheme that minimises redundancy among training samples to improve model generalisation and overall counting performance is developed.
People, Penguins and Petri Dishes: Adapting Object Counting Models to New Visual Domains and Object Types Without Forgetting
A technique to adapt a convolutional neural network (CNN) based object counter to additional visual domains and object types while still preserving the original counting function is proposed.
ResnetCrowd: A residual deep learning architecture for crowd counting, violent behaviour detection and crowd density level classification
Experiments show that a multi-task approach boosts individual task performance for all tasks and most notably for violent behaviour detection which receives a 9% boost in ROC curve AUC (Area under the curve).
Rules-By-Example - A Novel Approach to Semantic Indexing and Querying of Images
This paper proposes a new hybrid, user-assisted approach to semantic indexing of images, Rules-By-Example (RBE), which is based on a combination of RuleML and Query- by-Example, and believes that this approach, in combination with traditional solutions, will enable faster, more flexible, cost-effective and accurate semantic indexed images.
"Colearning" - Collaborative Open Learning through OER and Social Media
This chapter introduces the concept of coLearning as well as discussing how open learning networks can produce, share and reuse OER collaboratively through social media. COLEARNING OBJECTIVES
Holistic features for real-time crowd behaviour anomaly detection
A new approach to crowd behaviour anomaly detection that uses a set of efficiently computed, easily interpretable, scene-level holistic features that combines two features from the literature: crowd collectiveness and crowd conflict, with two newly developed crowd features: mean motion speed and a new formulation of crowd density is presented.
Dynamic Generation of Intelligent Multimedia Presentations through Semantic Inferencing
A system, based on this architecture, which was developed as a service to run over OAI archives - but is applicable to any repositories containing mixed-media resources described using Dublin Core, and which may be able to generate new knowledge by exposing previously unrecognized connections.
MPEG-7 based Multimedia Ontologies: Interoperability Support or Interoperability Issue?
Four independent approaches to build a multimedia ontology based on the MPEG-7 standard are described and the similarities and differences between them are discussed.
Building and Indexing a Distributed Multimedia Presentation Archive Using SMIL
This paper proposes an approach to the problem of generating metadata for composite mixed-media digital objects by appropriately combining and exploiting existing knowledge or metadata associated
Supporting the Collaborative Adaptation of Open Educational Resources: The OpenScout Tool Library
Open educational resources (OER) have recently emerged as an answer to the need for open and reusable educational material, freely available online. Educational institutions, private organisations,