This thesis investigates cooperative and intelligent control of autonomous multi-robot systems in a dynamic, unstructured and unknown environment and makes significant original contributions with regard to self-deterministic learning for robot cooperation, evolutionary optimization of robotic actions, improvement of system robustness, vision-based object tracking, and real-time performance. A distributed multi-robot architecture is developed which will facilitate operation of a cooperative multi-robot system in a dynamic and unknown environment in a self-improving, robust, and real-time manner. It is a fully distributed and hierarchical architecture with three levels. By combining several popular AI, soft computing, and control techniques such as learning, planning, reactive paradigm, optimization, and hybrid control, the developed architecture is expected to facilitate effective autonomous operation of cooperative multi-robot systems in a dynamically changing, unknown, and unstructured environment. A machine learning technique is incorporated into the developed multi-robot system for self-deterministic and self-improving cooperation and coping with uncertainties in the environment. A modified Q-learning algorithm termed Sequential Q-learning with Kalman Filtering (SQKF) is developed in the thesis, which can provide fast multi-robot learning. By arranging the robots to learn according to a predefined sequence, modeling the effect of the actions of other robots in the work environment as Gaussian white noise and estimating this noise online with a Kalman filter, the SQKF algorithm seeks to solve several key problems in multi-robot learning. As a part of low-level sensing and control in the proposed multi-robot architecture, a fast computer vision algorithm for color-blob tracking is developed to track multiple moving objects in the environment. By removing the brightness and saturation information in an image and filtering unrelated information based on statistical features and domain knowledge, the algorithm solves the problems of uneven illumination in the environment and improves real-time performance.

No description available.

No description available.

No description available.

This thesis investigates cooperative and intelligent control of autonomous multi-robot systems in a dynamic, unstructured and unknown environment and makes significant original contributions with regard to self-deterministic learning for robot cooperation, evolutionary optimization of robotic actions, improvement of system robustness, vision-based object tracking, and real-time performance. A distributed multi-robot architecture is developed which will facilitate operation of a cooperative multi-robot system in a dynamic and unknown environment in a self-improving, robust, and real-time manner. It is a fully distributed and hierarchical architecture with three levels. By combining several popular AI, soft computing, and control techniques such as learning, planning, reactive paradigm, optimization, and hybrid control, the developed architecture is expected to facilitate effective autonomous operation of cooperative multi-robot systems in a dynamically changing, unknown, and unstructured environment. A machine learning technique is incorporated into the developed multi-robot system for self-deterministic and self-improving cooperation and coping with uncertainties in the environment. A modified Q-learning algorithm termed Sequential Q-learning with Kalman Filtering (SQKF) is developed in the thesis, which can provide fast multi-robot learning. By arranging the robots to learn according to a predefined sequence, modeling the effect of the actions of other robots in the work environment as Gaussian white noise and estimating this noise online with a Kalman filter, the SQKF algorithm seeks to solve several key problems in multi-robot learning. As a part of low-level sensing and control in the proposed multi-robot architecture, a fast computer vision algorithm for color-blob tracking is developed to track multiple moving objects in the environment. By removing the brightness and saturation information in an image and filtering unrelated information based on statistical features and domain knowledge, the algorithm solves the problems of uneven illumination in the environment and improves real-time performance.

PhD / The aim of this thesis was to investigate the possibility of integrating frozen-thawed boar semen into reproductive technologies and into commercial production of pigs in Australia. This was to be achieved by establishing a semen freezing and AI regime that was of a standard acceptable to industry, and integrating the resultant frozen-thawed sperm into other reproductive technologies, such as flow cytometric sperm sorting and IVF. Initially, a protocol for freezing and thawing boar semen was established, based on the method described by Westendorf et al. (1975) and attempts were made to modify this protocol to improve the post-thaw sperm quality, as determined by in vitro assessment of motility, acrosome integrity and longevity. First, the egg yolk used in the freezing extenders was investigated, and the chicken yolk was replaced with either duck or quail yolk. It was shown that there was no benefit in substituting yolk from duck or quail for the chicken yolk traditionally used in freezing extender. Second, the effect of seminal plasma addition to the freezing extender, or seminal plasma addition to resuspension medium post-thaw was tested. Incorporating whole seminal plasma into the freezing extender at levels above 50% was found to be detrimental to post-thaw sperm quality. Reducing levels to 20% of the final volume improved acrosome integrity, but adversely affected motility of sperm. However, adding 20% seminal plasma to the resuspension medium used after thawing of boar semen had no significant influence on sperm quality compared with resuspension in medium without seminal plasma. The antioxidant catalase, and the iron chelator desferal added to the freezing extender, did not improve post-thaw sperm quality, nor was any benefit seen with addition of these substrates to the resuspension medium post-thaw. However, the bioactive phospholipid PAF and its regulating enzyme PAF:AH appeared to enhance post-thaw motility and acrosome integrity of sperm, respectively, when added to the semen pre-freezing. Unfortunately, due to the restrictions imposed on rPAF:AH as a research drug, it was not possible to test the in vivo effects at this time. After the in vitro experiments were completed, the in vivo fertility of frozen-thawed sperm was tested using the optimal freezing protocol and a novel technology, enabling non-surgical deep intrauterine insemination of sows. The aim was to establish the lowest possible dose of frozen-thawed sperm that could be used, without compromising fertility. Successful pregnancies were achieved with doses as low as 62.5 x 106 frozen-thawed sperm but the farrowing rates were too low to be practicable on a commercial scale. This is the first report of litters born after insemination of such a low dose of frozen-thawed sperm and using the novel DIU insemination technique. However, it was concluded that a double dose of 250 x 106 frozen-thawed sperm was the minimum dose required for maintaining acceptable fertility. Reduction in sperm numbers to such an extent made it possible to consider non-surgical insemination of sex-sorted, frozen-thawed semen. Previously, pregnancies had been achieved only after surgical insemination of sex-sorted boar sperm, or with DIU insemination of unfrozen sperm, immediately after sex-sorting. The low numbers of sex-sorted sperm available restricted the inseminate dose used here to 50 x106 motile sperm. A litter of 5 piglets was born after a low-dose, DIU insemination of sex-sorted, frozen-thawed sperm. This is the first report of piglets born after insemination with sex-sorted frozen-thawed sperm and non-surgical insemination. The low farrowing rate achieved in this experiment prompted the investigation of integrating sex-sorted, frozen-thawed boar sperm into IVF. Morulae were produced after IVF with sex-sorted, frozen-thawed sperm and successfully transferred using non-surgical techniques. This is the first report of pregnancy achieved with non-surgical transfer of embryos produced after IVF and IVC of IVM oocytes with sex-sorted, frozen-thawed boar sperm. Unfortunately, the pregnancy did not hold, and the embryos were lost prior to Day 32, but PCR of non-transferred embryos confirmed successful pre-selection of sex. Overall, this thesis demonstrated that it is still not economically feasible to incorporate frozen-thawed boar semen into the commercial production of pigs although it has considerable application in breeding programmes. However, the development of novel techniques enabling reduction in sperm dose, and for non-surgical transfer of embryos into recipient sows and incorporation of frozen-thawed semen into these technologies means that progress is being made with the integration of reproductive technologies and frozen-thawed semen into the pig industry.

A fundamental question in visual neuroscience is how to represent image structure. The most common representational schemes rely on differential operators that compare adjacent image regions. While well-suited to encoding local relationships, such operators have significant drawbacks. Specifically, each filter’s span is confounded with the size of its sub-fields, making it difficult to compare small regions across large distances. We find that such long-distance comparisons are more tolerant to common image transformations than purely local ones, suggesting they may provide a useful vocabulary for image encoding. .We introduce the “Dissociated Dipole,” or “Sticks” operator, for encoding non-local image relationships. This operator de-couples filter span from sub-field size, enabling parametric movement between edge and region-based representation modes. We report on the perceptual plausibility of the operator, and the computational advantages of non-local encoding. Our results suggest that non-local encoding may be an effective scheme for representing image structure.

We introduce and explore an approach to estimating statisticalsignificance of classification accuracy, which is particularly usefulin scientific applications of machine learning where highdimensionality of the data and the small number of training examplesrender most standard convergence bounds too loose to yield ameaningful guarantee of the generalization ability of theclassifier. Instead, we estimate statistical significance of theobserved classification accuracy, or the likelihood of observing suchaccuracy by chance due to spurious correlations of thehigh-dimensional data patterns with the class labels in the giventraining set. We adopt permutation testing, a non-parametric techniquepreviously developed in classical statistics for hypothesis testing inthe generative setting (i.e., comparing two probabilitydistributions). We demonstrate the method on real examples fromneuroimaging studies and DNA microarray analysis and suggest atheoretical analysis of the procedure that relates the asymptoticbehavior of the test to the existing convergence bounds.

Alternative pre-messenger RNA splicing affects a majority of human genes and plays important roles in development and disease. Alternative splicing (AS) events conserved since the divergence of human and mouse are likely of primary biological importance, but relatively few such events are known. Here we describe sequence features that distinguish exons subject to evolutionarily conserved AS, which we call 'alternative-conserved exons' (ACEs) from other orthologous human/mouse exons, and integrate these features into an exon classification algorithm, ACEScan. Genome-wide analysis of annotated orthologous human-mouse exon pairs identified ~2,000 predicted ACEs. Alternative splicing was verified in both human and mouse tissues using an RT-PCR-sequencing protocol for 21 of 30 (70%) predicted ACEs tested, supporting the validity of a majority of ACEScan predictions. By contrast, AS was observed in mouse tissues for only 2 of 15 (13%) tested exons which had EST or cDNA evidence of AS in human but were not predicted ACEs, and was never observed for eleven negative control exons in human or mouse tissues. Predicted ACEs were much more likely to preserve reading frame, and less likely to disrupt protein domains than other AS events, and were enriched in genes expressed in the brain and in genes involved in transcriptional regulation, RNA processing and development. Our results also imply that the vast majority of AS events represented in the human EST databases are not conserved in mouse, and therefore may represent aberrant, disease- or allele-specific, or highly lineage-restricted splicing events.

Discriminative learning is challenging when examples are setsof local image features, and the sets vary in cardinality and lackany sort of meaningful ordering. Kernel-based classificationmethods can learn complex decision boundaries, but a kernelsimilarity measure for unordered set inputs must somehow solve forcorrespondences -- generally a computationally expensive task thatbecomes impractical for large set sizes. We present a new fastkernel function which maps unordered feature sets tomulti-resolution histograms and computes a weighted histogramintersection in this space. This ``pyramid match" computation islinear in the number of features, and it implicitly findscorrespondences based on the finest resolution histogram cell wherea matched pair first appears. Since the kernel does not penalize thepresence of extra features, it is robust to clutter. We show thekernel function is positive-definite, making it valid for use inlearning algorithms whose optimal solutions are guaranteed only forMercer kernels. We demonstrate our algorithm on object recognitiontasks and show it to be dramatically faster than currentapproaches.