Semantic change detection through large-scale learning. This project aims to develop technologies which understand the content of images before higher-level analysis is performed. This approach is intended to allow more accurate and reliable decisions to be made using automated image analysis than has previously been possible. The project will particularly investigate the detection of change in the contents of an image.
Visual tracking of multiple objects: A stochastic geometrical approach. Reliable real-time visual multiple-object tracking techniques will open up new applications that enhance the quality of life such as driving safety, traffic monitoring, home security, security and surveillance of public facilities. These new applications have huge commercial potentials, and the technology developed from this project will provide the catalyst for vibrant new industries to grow. In addition, the expertise deve ....Visual tracking of multiple objects: A stochastic geometrical approach. Reliable real-time visual multiple-object tracking techniques will open up new applications that enhance the quality of life such as driving safety, traffic monitoring, home security, security and surveillance of public facilities. These new applications have huge commercial potentials, and the technology developed from this project will provide the catalyst for vibrant new industries to grow. In addition, the expertise developed from the project will provide a competitive edge for Australian industries in aerospace, oceanography, robotics, remote sensing, and biomedical engineering. Read moreRead less
Online Learning for Large Scale Structured Data in Complex Situations. Online Learning (OL) is the process of predicting answers for a sequence of questions. OL has enjoyed much attention in recent years due to its natural ability of processing large scale non-structured data and adapting to a changing environment. However, OL has three weaknesses: it does not scale for structured data; it often assumes that all of the data are equally important; it often considers that all of the data are compl ....Online Learning for Large Scale Structured Data in Complex Situations. Online Learning (OL) is the process of predicting answers for a sequence of questions. OL has enjoyed much attention in recent years due to its natural ability of processing large scale non-structured data and adapting to a changing environment. However, OL has three weaknesses: it does not scale for structured data; it often assumes that all of the data are equally important; it often considers that all of the data are complete and noise-free. These weaknesses limit its utility, because real data such as those that must be analysed in processing social networks, fraud detection do not satisfy the restrictions. The aim of this project is to develop theoretical and practical advances in OL that overcome the existing weaknesses.Read moreRead less
Probabilistic Graphical Models For Interventional Queries. The project intends to develop methods to suggest how to optimally intervene so that the future state of the system will best suit our interests. The power of probabilistic graphical models to model complex relationships and interactions among a large number of variables facilitates many applications. However, such models only aim to understand the underlying environment. What is ultimately needed in many real-world applications is to su ....Probabilistic Graphical Models For Interventional Queries. The project intends to develop methods to suggest how to optimally intervene so that the future state of the system will best suit our interests. The power of probabilistic graphical models to model complex relationships and interactions among a large number of variables facilitates many applications. However, such models only aim to understand the underlying environment. What is ultimately needed in many real-world applications is to suggest how we ought to intervene or act, so as to alter the environment to best suit our interests. The proposed project aims to achieve this using probabilistic graphical models on massive real-world data sets, thus facilitating a variety of applications from health care to commerce and the environment.Read moreRead less
Statistical Methods of Model Fitting and Segmentation in Computer Vision. Electronic sensors such as cameras and lasers can provide a rich source of information about the position, shape, and motion of objects around us. However, to extract this information in a reliable, automatic, and accurate way requires a sophisticated statistical theory of the process. Example applications include: video surveillance (better automatic detection of moving people and vehicles and of characterising what those ....Statistical Methods of Model Fitting and Segmentation in Computer Vision. Electronic sensors such as cameras and lasers can provide a rich source of information about the position, shape, and motion of objects around us. However, to extract this information in a reliable, automatic, and accurate way requires a sophisticated statistical theory of the process. Example applications include: video surveillance (better automatic detection of moving people and vehicles and of characterising what those people and vehicles are doing), industrial prototyping and inspection (measuring the size and shape of objects), urban planning (laser scanning streetscapes to create computer models of cities), entertainment industry (movie special effects and games), etc. Read moreRead less
Learning to see in 3D. The project aims to endow machine vision with an ability we, as humans, use almost constantly: to judge 3D properties from a 2D image. This extremely useful ability will be applied to digital images to obtain 3D measurements and aid in automating tasks such as mining, surveying, medical diagnosis, and visual effects in movies.
Recognising and reconstructing objects in real time from a moving camera. This project will use a moving camera to estimate the three-dimensional shape and identity of objects and surfaces it can see. This ability, which we humans use all the time, has wide application in automation including driver assistance, exploring hazardous environments, robotics, remote collaboration, and the creation of three-dimensional models for entertainment.
Optimal Robust Fitting under the Framework of LP-Type Problems. The project aims to develop algorithms to support the development of robust and accurate computer vision systems. Real-world visual data (images, videos) is inherently noisy and outlier prone. To build computer vision systems that work reliably in the real world, it is necessary to ensure that the underlying algorithms are robust and efficient. The project aims to devise novel algorithms that can compute the best possible result giv ....Optimal Robust Fitting under the Framework of LP-Type Problems. The project aims to develop algorithms to support the development of robust and accurate computer vision systems. Real-world visual data (images, videos) is inherently noisy and outlier prone. To build computer vision systems that work reliably in the real world, it is necessary to ensure that the underlying algorithms are robust and efficient. The project aims to devise novel algorithms that can compute the best possible result given the input data in a short amount of time. The expected outcomes would support the construction of reliable and accurate computer vision-based systems, such as large-scale 3-D reconstruction from photo collections, self-driving cars and domestic robots.Read moreRead less
Australian Laureate Fellowships - Grant ID: FL130100102
Funder
Australian Research Council
Funding Amount
$3,179,946.00
Summary
Lifelong computer vision systems. This project will create a computer vision system that can produce a detailed environmental map in real time, turning standard video cameras into sensors that 'understand' a scene with basic semantic tools. This high-level sensing will unlock a wide range of applications for autonomous systems.
Whole image understanding by convolutions on graphs. This project seeks to develop technologies that will help computer vision interpret the whole visible scene, rather than just some of the objects therein. Existing automated methods for understanding images perform well at recognising specific objects in canonical poses, but the problem of whole image interpretation is far more challenging. Convolutional neural networks (CNN) have underpinned recent progress in object recognition, but whole-im ....Whole image understanding by convolutions on graphs. This project seeks to develop technologies that will help computer vision interpret the whole visible scene, rather than just some of the objects therein. Existing automated methods for understanding images perform well at recognising specific objects in canonical poses, but the problem of whole image interpretation is far more challenging. Convolutional neural networks (CNN) have underpinned recent progress in object recognition, but whole-image understanding cannot be tackled similarly because the number of possible combinations of objects is too large. The project thus proposes a graph-based generalisation of the CNN approach which allows scene structure to be learned explicitly. This would represent an important step towards providing computers with robust vision, allowing them to interact with their environment.Read moreRead less