Design expressions for the in-plane shear capacity of confined masonry shear walls containing squat panels

In-plane shear capacity formulation of reinforced masonry is commonly conceived as the sum of the capacities of three parameters, viz, the masonry, the reinforcement, and the precompression. The term “masonry” incorporates the aspect ratio of the wall without any regard to the aspect ratio of the panels inscribed (and hence confined) by the vertical and the horizontal reinforced grout cores. This paper proposes design expressions in which the aspect ratio of such panels is explicitly included. For this purpose, the grouted confining cores are regarded as a grid of confining elements within which the panels are positioned. These confined masonry panels are then considered as building blocks for multi-bay, multi-storied confined masonry shear walls and analyzed using an experimentally validated macroscopic finite-element model. Results of the analyzes of 161 confined masonry walls containing panels of height to length ratio less than 1.0 have been regressed to formulate design expressions. These expressions have been first validated using independent test data sets and then compared with the existing equations in some selected international design standards. The concept of including the unreinforced masonry panel aspect ratio as an additional term in the design expression for partially grouted/confined masonry shear walls is recommended based on the conclusions from this paper.

Activity recognition using binary tree SVM

This paper presents an effective classification method based on Support Vector Machines (SVM) in the context of activity recognition. Local features that capture both spatial and temporal information in activity videos have made significant progress recently. Efficient and effective features, feature representation and classification plays a crucial role in activity recognition. For classification, SVMs are popularly used because of their simplicity and efficiency; however the common multi-class SVM approaches applied suffer from limitations including having easily confused classes and been computationally inefficient. We propose using a binary tree SVM to address the shortcomings of multi-class SVMs in activity recognition. We proposed constructing a binary tree using Gaussian Mixture Models (GMM), where activities are repeatedly allocated to subnodes until every new created node contains only one activity. Then, for each internal node a separate SVM is learned to classify activities, which significantly reduces the training time and increases the speed of testing compared to popular the `one-against-the-rest' multi-class SVM classifier. Experiments carried out on the challenging and complex Hollywood dataset demonstrates comparable performance over the baseline bag-of-features method.

Improving daylighting performance of mirrored light pipes

This paper assesses and compares the performances of two daylight collection strategies, one passive and one active, for large-scale mirrored light pipes (MLP) illuminating deep plan buildings. Both strategies use laser cut panels (LCP) as the main component of the collection system. The passive system comprises LCPs in pyramid form, whereas the active system uses a tiled LCP on a simple rotation mechanism that rotates 360° in 24 hours. Performance is assessed using scale model testing under sunny sky conditions and mathematical modelling. Results show average illuminance levels for the pyramid LCP ranging from 50 to 250 lux and 150 to 200 lux for the rotating LCPs. Both systems improve the performance of a MLP. The pyramid LCP increases the performance of a MLP by 2.5 times and the rotating LCP by 5 times, when compared to an open pipe particularly for low sun elevation angles.