An infinite-horizon robust filter for uncertain hidden Markov Models with conditional relative entropy constraints

We consider a robust filtering problem for uncertain discrete-time, homogeneous, first-order, finite-state hidden Markov models (HMMs). The class of uncertain HMMs considered is described by a conditional relative entropy constraint on measures perturbed from a nominal regular conditional probability distribution given the previous posterior state distribution and the latest measurement. Under this class of perturbations, a robust infinite horizon filtering problem is first formulated as a constrained optimization problem before being transformed via variational results into an unconstrained optimization problem; the latter can be elegantly solved using a risk-sensitive information-state based filtering.

Anomalous event detection using a semi-two dimensional hidden Markov model

The rapid increase in the deployment of CCTV systems has led to a greater demand for algorithms that are able to process incoming video feeds. These algorithms are designed to extract information of interest for human operators. During the past several years, there has been a large effort to detect abnormal activities through computer vision techniques. Typically, the problem is formulated as a novelty detection task where the system is trained on normal data and is required to detect events which do not fit the learned `normal' model. Many researchers have tried various sets of features to train different learning models to detect abnormal behaviour in video footage. In this work we propose using a Semi-2D Hidden Markov Model (HMM) to model the normal activities of people. The outliers of the model with insufficient likelihood are identified as abnormal activities. Our Semi-2D HMM is designed to model both the temporal and spatial causalities of the crowd behaviour by assuming the current state of the Hidden Markov Model depends not only on the previous state in the temporal direction, but also on the previous states of the adjacent spatial locations. Two different HMMs are trained to model both the vertical and horizontal spatial causal information. Location features, flow features and optical flow textures are used as the features for the model. The proposed approach is evaluated using the publicly available UCSD datasets and we demonstrate improved performance compared to other state of the art methods.

Hidden Innovation : Policy, Industry and the Creative Sector

The latest book in the internationally acclaimed Creative Economy series. The term ‘two cultures’ was coined more than 50 years ago by scientist and novelist C.P. Snow to describe the divergence in the world views and methods of scientists and the creative sector. This divergence has meant that innovation systems and policies have focussed for decades on science, engineering, technology and medicine and the industries that depend on them. The humanities, arts and social sciences have been bitt players at best; their contributions hidden from research agendas, policy and program initiatives, and the public mind. But structural changes to advanced economies and societies have brought services industries and the creative sector to greater prominence as key contributors to innovation. Hidden Innovation peels back the veil, tracing the way innovation occurs through new forms of screen production enabled by social media platforms as well as in public broadcasting. It shows that creative workers are contributing fresh ideas across the economy, and how creative cities debates need reframing. It traces how policies globally are beginning to catch up with the changing social and economic realities. In his new book, Cunningham argues that the innovation framework offers the best opportunity in decades to reassess and refresh the case for the public role of the humanities, particularly the media, cultural and communication studies disciplines.

Social signal processing for pain monitoring using a hidden conditional random field

utomatic pain monitoring has the potential to greatly improve patient diagnosis and outcomes by providing a continuous objective measure. One of the most promising methods is to do this via automatically detecting facial expressions. However, current approaches have failed due to their inability to: 1) integrate the rigid and non-rigid head motion into a single feature representation, and 2) incorporate the salient temporal patterns into the classification stage. In this paper, we tackle the first problem by developing a “histogram of facial action units” representation using Active Appearance Model (AAM) face features, and then utilize a Hidden Conditional Random Field (HCRF) to overcome the second issue. We show that both of these methods improve the performance on the task of pain detection in sequence level compared to current state-of-the-art-methods on the UNBC-McMaster Shoulder Pain Archive.

Hollywood's hidden advantage : creative suits

Creativity in Hollywood is not just about telling stories onscreen. Deal making is the name of the game in Hollywood from globally released franchised blockbusters to art house releases. Riding the currents of the twentieth century Hollywood has maintained dominance with its highly diversified production slate built on creative financing solutions. Using historical and recent case studies, the presentation will look behind the images at the numbers and discuss how 'the suits' have been, and continue to be just as creative as the 'creatives' in contemporary Hollywood.

Male reproductive health disorders among Aboriginal and Torres Strait Islander men : a hidden problem?

OBJECTIVE: To better understand help-seeking behaviours and reproductive health disorders among Aboriginal and Torres Strait Islander men. DESIGN, SETTING AND PARTICIPANTS: A cross-sectional mixed-methods study conducted from 1 May 2004 to 30 April 2005 of 293 Aboriginal and Torres Strait Islander men aged 18 years and over from urban, rural and remote communities in the Northern Territory and Queensland. MAIN OUTCOME MEASURES: Subscale of the International Index of Erectile Function, self-reported help-seeking behaviours for erectile dysfunction (ED) and prostate disease, thematic analysis of semi-structured interviews and focus groups. RESULTS: The prevalence of moderate-to-severe ED increased across age groups, from about 10% in younger men (under 35 years) to 28% in men aged 55-74 years. Moderate-to-severe ED was strongly associated with reporting a chronic condition (odds ratio [OR], 3.67) and residing in a remote area (OR, 2.94). Aboriginal and Torres Strait Islander men aged 40-59 years showed similar low levels of help-seeking behaviours compared with non-Indigenous men from a comparable population-based study. About half of the men with ED saw a doctor or received treatment for ED in each population. While prostate cancer rates were low in both studies, testing for prostate problems was less frequent in Aboriginal and Torres Strait Islander men (11.4%) than in non-Indigenous men (34.1%, P < 0.001), despite similar levels of concern about prostate cancer. Barriers to help-seeking included shame, culturally inappropriate services and lack of awareness. CONCLUSION: This study, the first to investigate reproductive health of Aboriginal and Torres Strait Islander men, found low levels of help-seeking behaviours for reproductive health disorders, with implications for missing a predictor of chronic disease and late diagnosis of prostate disease.

Segmentation of cone-beam CT using a hidden Markov random field with informative priors

Purpose: Flat-detector, cone-beam computed tomography (CBCT) has enormous potential to improve the accuracy of treatment delivery in image-guided radiotherapy (IGRT). To assist radiotherapists in interpreting these images, we use a Bayesian statistical model to label each voxel according to its tissue type. Methods: The rich sources of prior information in IGRT are incorporated into a hidden Markov random field (MRF) model of the 3D image lattice. Tissue densities in the reference CT scan are estimated using inverse regression and then rescaled to approximate the corresponding CBCT intensity values. The treatment planning contours are combined with published studies of physiological variability to produce a spatial prior distribution for changes in the size, shape and position of the tumour volume and organs at risk (OAR). The voxel labels are estimated using the iterated conditional modes (ICM) algorithm. Results: The accuracy of the method has been evaluated using 27 CBCT scans of an electron density phantom (CIRS, Inc. model 062). The mean voxel-wise misclassification rate was 6.2%, with Dice similarity coefficient of 0.73 for liver, muscle, breast and adipose tissue. Conclusions: By incorporating prior information, we are able to successfully segment CBCT images. This could be a viable approach for automated, online image analysis in radiotherapy.

Using images to communicate the hidden struggles of life on dialysis

Background: Nurse-patient communication in the hemodialysis context is unique given the amount of time spent together in a confined clinical room. Poor communication may lead to low quality nursing care and undesirable patient outcomes. Aim: To explore the use of images as a visual communication technique for nurses and patients in the hemodialysis context. Methods: Descriptive qualitative design. Fifty two cards containing specific photos, illustrations and words were used in conversations between patients (n = 9) and one of two nurse interviewers about being on hemodialysis. Interview transcripts were thematically analysed. Findings: An overall theme titled ‘revealing the hidden struggles of living on dialysis’ conceptually captured three sub-themes: (1) the increased importance of relationships; (2) the struggle with money; and (3) quality over quantity of life. The cards assisted in uncovering these often covert (to nurses) aspects of dialysis patients’ lives. Conclusion: Nurses may need to be aware of the dialysis patients’ hidden struggles which include the importance of relationships, financial issues and the importance of quality aspects such as travel. The use of images may assist in revealing the important issues for each patient struggling with the restrictive life that is imposed by dialysis.

Segmentation of cone-beam CT using a hidden Markov random field with informative priors

Cone-beam computed tomography (CBCT) has enormous potential to improve the accuracy of treatment delivery in image-guided radiotherapy (IGRT). To assist radiotherapists in interpreting these images, we use a Bayesian statistical model to label each voxel according to its tissue type. The rich sources of prior information in IGRT are incorporated into a hidden Markov random field model of the 3D image lattice. Tissue densities in the reference CT scan are estimated using inverse regression and then rescaled to approximate the corresponding CBCT intensity values. The treatment planning contours are combined with published studies of physiological variability to produce a spatial prior distribution for changes in the size, shape and position of the tumour volume and organs at risk. The voxel labels are estimated using iterated conditional modes. The accuracy of the method has been evaluated using 27 CBCT scans of an electron density phantom. The mean voxel-wise misclassification rate was 6.2\%, with Dice similarity coefficient of 0.73 for liver, muscle, breast and adipose tissue. By incorporating prior information, we are able to successfully segment CBCT images. This could be a viable approach for automated, online image analysis in radiotherapy.