A Generic Deep-Learning-Based Approach for Automated Surface Inspection.

Automated surface inspection (ASI) is a challenging task in industry, as collecting training dataset is usually costly and related methods are highly dataset-dependent. In this paper, a generic approach that requires small training data for ASI is proposed. First, this approach builds classifier on the features of image patches, where the features are transferred from a pretrained deep learning network. Next, pixel-wise prediction is obtained by convolving the trained classifier over input image. An experiment on three public and one industrial data set is carried out. The experiment involves two tasks: 1) image classification and 2) defect segmentation. The results of proposed algorithm are compared against several best benchmarks in literature. In the classification tasks, the proposed method improves accuracy by 0.66%-25.50%. In the segmentation tasks, the proposed method reduces error escape rates by 6.00%-19.00% in three defect types and improves accuracies by 2.29%-9.86% in all seven defect types. In addition, the proposed method achieves 0.0% error escape rate in the segmentation task of industrial data.

The emergence of urban land use patterns driven by dispersion and aggregation mechanisms.

We employ a cellular-automata to reconstruct the land use patterns of cities that we characterize by two measures of spatial heterogeneity: (a) a variant of spatial entropy, which measures the spread of residential, business, and industrial activity sectors, and (b) an index of dissimilarity, which quantifies the degree of spatial mixing of these land use activity parcels. A minimalist and bottom-up approach is adopted that utilizes a limited set of three parameters which represent the forces which determine the extent to which each of these sectors spatially aggregate into clusters. The dispersion degrees of the land uses are governed by a fixed pre-specified power-law distribution based on empirical observations in other cities. Our method is then used to reconstruct land use patterns for the city state of Singapore and a selection of North American cities. We demonstrate the emergence of land use patterns that exhibit comparable visual features to the actual city maps defining our case studies whilst sharing similar spatial characteristics. Our work provides a complementary approach to other measures of urban spatial structure that differentiate cities by their land use patterns resulting from bottom-up dispersion and aggregation processes.

Evaluating temporal factors in combined interventions of workforce shift and school closure for mitigating the spread of influenza.

BACKGROUND: It is believed that combined interventions may be more effective than individual interventions in mitigating epidemic. However there is a lack of quantitative studies on performance of the combination of individual interventions under different temporal settings. METHODOLOGY/PRINCIPAL FINDINGS: To better understand the problem, we develop an individual-based simulation model running on top of contact networks based on real-life contact data in Singapore. We model and evaluate the spread of influenza epidemic with intervention strategies of workforce shift and its combination with school closure, and examine the impacts of temporal factors, namely the trigger threshold and the duration of an intervention. By comparing simulation results for intervention scenarios with different temporal factors, we find that combined interventions do not always outperform individual interventions and are more effective only when the duration is longer than 6 weeks or school closure is triggered at the 5% threshold; combined interventions may be more effective if school closure starts first when the duration is less than 4 weeks or workforce shift starts first when the duration is longer than 4 weeks. CONCLUSIONS/SIGNIFICANCE: We therefore conclude that identifying the appropriate timing configuration is crucial for achieving optimal or near optimal performance in mitigating the spread of influenza epidemic. The results of this study are useful to policy makers in deliberating and planning individual and combined interventions.

Exploratory analysis of cell-based screening data for phenotype identification in drug-siRNA study.

Most phenotype-identification methods in cell-based screening assume prior knowledge about expected phenotypes or involve intricate parameter-setting. They are useful for analysis targeting known phenotype properties; but need exists to explore, with minimum presumptions, the potentially-interesting phenotypes derivable from data. We present a method for this exploration, using clustering to eliminate phenotype-labelling requirement and GUI visualisation to facilitate parameter-setting. The steps are: outlier-removal, cell clustering and interactive visualisation for phenotypes refinement. For drug-siRNA study, we introduce an auto-merging procedure to reduce phenotype redundancy. We validated the method on two Golgi apparatus screens and showcase its contribution for better understanding of screening-images.

Temporal factors in school closure policy for mitigating the spread of influenza.

Is school closure effective in mitigating influenza outbreaks? For Singapore, we developed an individual-based simulation model using real-life contact data. We evaluated the impacts of temporal factors - trigger threshold and duration - on the effectiveness of school closure as a mitigation policy. We found an upper bound of the duration of school closure, where further extension beyond which will not bring additional benefits to suppressing the attack rate and peak incidence. For school closure with a relatively short duration (< 6 weeks), it is more effective to start closure after a relatively longer delay from the first day of infection; if the duration of school closure is long (>6 weeks), however, it is better to start it as early as reasonable. Our studies reveal the critical importance of timing in school closure, especially in cost-cautious situations. Our studies also demonstrate the great potential of a properly developed individual-based simulation model in evaluating various disease control policies.