权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

AUTOMAC: AUTOmated Mouse behAviour reCognition

AUTOMAC：自动化鼠标行为识别

基本信息

批准号：
EP/N011074/1
负责人：
HUIYU ZHOU
金额：
$ 12.6万
依托单位：
Queen's University Belfast
依托单位国家：
英国
项目类别：
Research Grant
财政年份：
2016
资助国家：
英国
起止时间：
2016 至无数据
项目状态：
已结题

来源：
https://gtr.ukri.org/projects?ref=EP%2FN011074%2F1
关键词：
AUTOMAC AUTOmated Mouse behAviour reCognition

项目摘要

Context of the researchNeurodegenerative diseases are characterised by motor deficiencies. For many of them, there are no successful neuroprotective or neuroregenerative therapies clinically available. In order to address this problem, the development of valid animal models for motor disorders has become active growing and vibrant field in preclinical research. Behaviour analysis of laboratory animals has been recognised as a useful tool to assess therapeutic efficacy. The entire process consists of animal tracking and motion categorisation. Despite tremendous efforts made within the research community, there is no system which can perform reliable recognition of complex animal behaviours and interactions. In this project, a fully automated and trainable computer vision system is proposed to monitor and analyse complex mouse behaviours and interactions using video data recorded by calibrated cameras.Aims and objectivesScientific: (1) To develop a system of combining multi-camera tracking and a Hidden Markov Model.(2) To improve the multi-camera tracking performance combining covariance descriptors and probabilistic data association.(3) To accelerate K-means clustering using an approximate nearest neighbour algorithm.Practical:(1) To improve the performance and scalability of the existing behaviour analysis systems.(2) To widen the scope of the applicability of the developed tracker and the behaviour recognition system. (3) To associate healthcare applications with the image and vision computing community.Potential applications and benefits(1) This project will help researchers in the healthcare/medical community to significantly reduce annotation time/errors and hence improve medical research quality. The research outcomes of the proposed multidisciplinary project can reach both ICT and healthcare communities by our attendance at conferences in different domains. In the meantime, the research communities will benefit from our publications in journals and publicly accessible tools/databases for sharing skills and experiences.(2) The proposed research may be commercialised in the form of software tools. The UK hosts many companies that offer services related to the treatment of neurodegenerative diseases, e.g. GSK, and Orion pharmaceutical companies. These companies and their clients stand to profit from improvements in disease modelling and diagnosis/treatment techniques that depend on animal modelling using a system like that proposed in this project. Significant scientific improvements in this field will have a transformative effect on these businesses.(3) The technologies developed in this project can be directly transferred and applied in physical security, human computer interface and virtual reality. The tools developed in the proposed research can be used to monitor moving objects (e.g. humans and animals) in different set-ups (e.g. authentic and virtual environments). The software package produced in the proposed research can easily find its customers in manufacture and design, basic science, communication engineering, media and entertainment. As a result, there is great potential for wealth creation and boosted economic prosperity from the developed software package for a wider range of applications.

研究背景神经退行性疾病以运动缺陷为特征。对于他们中的许多人来说，临床上还没有成功的神经保护或神经再生疗法。为了解决这一问题，发展有效的运动障碍动物模型已成为临床前研究的活跃、日益增长和充满活力的领域。实验动物的行为分析已被认为是评估治疗效果的有用工具。整个过程包括动物跟踪和运动分类。尽管研究界做出了巨大努力，但还没有一个系统可以对复杂的动物行为和相互作用进行可靠的识别。在本项目中，提出了一种全自动化和可训练的计算机视觉系统，利用校准摄像机记录的视频数据来监控和分析复杂的老鼠行为和交互。目的和目标科学：(1)开发一个结合多摄像头跟踪和隐马尔可夫模型的系统。(2)结合协方差描述子和概率数据关联来提高多摄像头跟踪性能。(3)使用近似最近邻算法加速K-Means聚类。实践：(1)提高现有行为分析系统的性能和可扩展性。(2)扩大所开发的跟踪器和行为识别系统的适用范围。(3)将医疗保健应用程序与图像和视觉计算社区相关联。潜在的应用程序和益处(1)该项目将帮助医疗保健/医疗社区的研究人员显著减少注释时间/错误，从而提高医学研究质量。通过我们参加不同领域的会议，拟议的多学科项目的研究成果可以到达ICT和医疗保健社区。与此同时，研究界将受益于我们在期刊上发表的文章和公开可访问的工具/数据库，以分享技能和经验。(2)拟议的研究可能会以软件工具的形式商业化。英国拥有许多提供与神经退行性疾病治疗相关的服务的公司，例如GSK和猎户座制药公司。这些公司及其客户将从疾病建模和诊断/治疗技术的改进中获利，这些改进依赖于使用本项目中提议的系统的动物建模。这一领域的重大科学进步将对这些企业产生革命性的影响。(3)本项目开发的技术可以直接转移和应用于物理安全、人机界面和虚拟现实。拟议研究中开发的工具可用于监控不同设置(例如真实和虚拟环境)中的移动对象(例如人和动物)。在本研究中产生的软件包可以很容易地在制造和设计、基础科学、通信工程、媒体和娱乐等领域找到客户。因此，为更广泛的应用开发的软件包具有创造财富和促进经济繁荣的巨大潜力。