Improving spoken term detection using complementary information

This research has made contributions to the area of spoken term detection (STD), defined as the process of finding all occurrences of a specified search term in a large collection of speech segments. The use of visual information in the form of lip movements of the speaker in addition to audio and the use of topic of the speech segments, and the expected frequency of words in the target speech domain, are proposed. By using these complementary information, improvement in the performance of STD has been achieved which enables efficient search of key words in large collection of multimedia documents.

Control of Ship Roll Motion

The undesirable effects of roll motion of ships (rocking about the longitudinal axis) became noticeable in the mid-nineteenth century when significant changes were introduced to the design of ships as a result of sails being replaced by steam engines and the arrangement being changed from broad to narrow hulls. The combination of these changes led to lower transverse stability (lower restoring moment for a given angle of roll) with the consequence of larger roll motion. The increase in roll motion and its effect on cargo and human performance lead to the development several control devices that aimed at reducing and controlling roll motion. The control devices most commonly used today are fin stabilizers, rudder, anti-roll tanks, and gyrostabilizers. The use of different types of actuators for control of ship roll motion has been amply demonstrated for over 100 years. Performance, however, can still fall short of expectations because of difficulties associated with control system design, which have proven to be far from trivial due to fundamental performance limitations and large variations of the spectral characteristics of wave-induced roll motion. This short article provides an overview of the fundamentals of control design for ship roll motion reduction. The overview is limited to the most common control devices.

Energy-based motion control of a slender hull unmanned underwater vehicle

This paper presents a motion control system for tracking of attitude and speed of an underactuated slender-hull unmanned underwater vehicle. The feedback control strategy is developed using the Port-Hamiltonian theory. By shaping of the target dynamics (desired dynamic response in closed loop) with particular attention to the target mass matrix, the influence of the unactuated dynamics on the controlled system is suppressed. This results in achievable dynamics independent of stable uncontrolled states. Throughout the design, the insight of the physical phenomena involved is used to propose the desired target dynamics. Integral action is added to the system for robustness and to reject steady disturbances. This is achieved via a change of coordinates that result in input-to-state stable (ISS) target dynamics. As a final step in the design, an anti-windup scheme is implemented to account for limited actuator capacity, namely saturation. The performance of the design is demonstrated through simulation with a high-fidelity model.

From vision to actions: Towards adaptive and autonomous humanoid robots

Although robotics research has seen advances over the last decades robots are still not in widespread use outside industrial applications. Yet a range of proposed scenarios have robots working together, helping and coexisting with humans in daily life. In all these a clear need to deal with a more unstructured, changing environment arises. I herein present a system that aims to overcome the limitations of highly complex robotic systems, in terms of autonomy and adaptation. The main focus of research is to investigate the use of visual feedback for improving reaching and grasping capabilities of complex robots. To facilitate this a combined integration of computer vision and machine learning techniques is employed. From a robot vision point of view the combination of domain knowledge from both imaging processing and machine learning techniques, can expand the capabilities of robots. I present a novel framework called Cartesian Genetic Programming for Image Processing (CGP-IP). CGP-IP can be trained to detect objects in the incoming camera streams and successfully demonstrated on many different problem domains. The approach requires only a few training images (it was tested with 5 to 10 images per experiment) is fast, scalable and robust yet requires very small training sets. Additionally, it can generate human readable programs that can be further customized and tuned. While CGP-IP is a supervised-learning technique, I show an integration on the iCub, that allows for the autonomous learning of object detection and identification. Finally this dissertation includes two proof-of-concepts that integrate the motion and action sides. First, reactive reaching and grasping is shown. It allows the robot to avoid obstacles detected in the visual stream, while reaching for the intended target object. Furthermore the integration enables us to use the robot in non-static environments, i.e. the reaching is adapted on-the- fly from the visual feedback received, e.g. when an obstacle is moved into the trajectory. The second integration highlights the capabilities of these frameworks, by improving the visual detection by performing object manipulation actions.

Fault tolerant vessel stabiliser control system

A vessel stabilizer control system includes a sensor fault detection means which senses the availability of sensing signals from a gyrostabilizer precession motion sensor and a vessel roll motion sensor. The control system controls the action of a gyro-actuator which is mechanically coupled to a gyrostabilizer. The benefit of employing fault sensing of the sensors providing the process control variables is that the sensed number of available process control variables (or sensors) can be used to activate a tiered system of control modes. Each tiered control mode is designed to utilize the available process control variables to ensure safe and effective operation of the gyrostabilizer that is tolerant of sensor faults and loss of power supply. A control mode selector is provided for selecting the appropriate control mode based on the number of available process control variables.

Ship Seakeeping Operability, Motion Control, and Autonomy - A Bayesian Perspective

Ship seakeeping operability refers to the quantification of motion performance in waves relative to mission requirements. This is used to make decisions about preferred vessel designs, but it can also be used as comprehensive assessment of the benefits of ship-motion-control systems. Traditionally, operability computation aggregates statistics of motion computed over over the envelope of likely environmental conditions in order to determine a coefficient in the range from 0 to 1 called operability. When used for assessment of motion-control systems, the increase of operability is taken as the key performance indicator. The operability coefficient is often given the interpretation of the percentage of time operable. This paper considers an alternative probabilistic approach to this traditional computation of operability. It characterises operability not as a number to which a frequency interpretation is attached, but as a hypothesis that a vessel will attain the desired performance in one mission considering the envelope of likely operational conditions. This enables the use of Bayesian theory to compute the probability of that this hypothesis is true conditional on data from simulations. Thus, the metric considered is the probability of operability. This formulation not only adheres to recent developments in reliability and risk analysis, but also allows incorporating into the analysis more accurate descriptions of ship-motion-control systems since the analysis is not limited to linear ship responses in the frequency domain. The paper also discusses an extension of the approach to the case of assessment of increased levels of autonomy for unmanned marine craft.

A Bayesian Framework for the Assessment of Vision-based Weed and Fruit Detection and Classification Algorithms

This paper proposes new metrics and a performance-assessment framework for vision-based weed and fruit detection and classification algorithms. In order to compare algorithms, and make a decision on which one to use fora particular application, it is necessary to take into account that the performance obtained in a series of tests is subject to uncertainty. Such characterisation of uncertainty seems not to be captured by the performance metrics currently reported in the literature. Therefore, we pose the problem as a general problem of scientific inference, which arises out of incomplete information, and propose as a metric of performance the(posterior) predictive probabilities that the algorithms will provide a correct outcome for target and background detection. We detail the framework through which these predicted probabilities can be obtained, which is Bayesian in nature. As an illustration example, we apply the framework to the assessment of performance of four algorithms that could potentially be used in the detection of capsicums (peppers).

On Visual Detection of Highly-occluded Objects for Harvesting Automation in Horticulture

Developing accurate and reliable crop detection algorithms is an important step for harvesting automation in horticulture. This paper presents a novel approach to visual detection of highly-occluded fruits. We use a conditional random field (CRF) on multi-spectral image data (colour and Near-Infrared Reflectance, NIR) to model two classes: crop and background. To describe these two classes, we explore a range of visual-texture features including local binary pattern, histogram of oriented gradients, and learn auto-encoder features. The pro-posed methods are evaluated using hand-labelled images from a dataset captured on a commercial capsicum farm. Experimental results are presented, and performance is evaluated in terms of the Area Under the Curve (AUC) of the precision-recall curves.Our current results achieve a maximum performance of 0.81AUC when combining all of the texture features in conjunction with colour information.

Nonparametric rank regression for analyzing water quality concentration data with multiple detection limits

Environmental data usually include measurements, such as water quality data, which fall below detection limits, because of limitations of the instruments or of certain analytical methods used. The fact that some responses are not detected needs to be properly taken into account in statistical analysis of such data. However, it is well-known that it is challenging to analyze a data set with detection limits, and we often have to rely on the traditional parametric methods or simple imputation methods. Distributional assumptions can lead to biased inference and justification of distributions is often not possible when the data are correlated and there is a large proportion of data below detection limits. The extent of bias is usually unknown. To draw valid conclusions and hence provide useful advice for environmental management authorities, it is essential to develop and apply an appropriate statistical methodology. This paper proposes rank-based procedures for analyzing non-normally distributed data collected at different sites over a period of time in the presence of multiple detection limits. To take account of temporal correlations within each site, we propose an optimal linear combination of estimating functions and apply the induced smoothing method to reduce the computational burden. Finally, we apply the proposed method to the water quality data collected at Susquehanna River Basin in United States of America, which dearly demonstrates the advantages of the rank regression models.

Statistical power calculation and sample size determination for environmental studies with data below detection limits

Power calculation and sample size determination are critical in designing environmental monitoring programs. The traditional approach based on comparing the mean values may become statistically inappropriate and even invalid when substantial proportions of the response values are below the detection limits or censored because strong distributional assumptions have to be made on the censored observations when implementing the traditional procedures. In this paper, we propose a quantile methodology that is robust to outliers and can also handle data with a substantial proportion of below-detection-limit observations without the need of imputing the censored values. As a demonstration, we applied the methods to a nutrient monitoring project, which is a part of the Perth Long-Term Ocean Outlet Monitoring Program. In this example, the sample size required by our quantile methodology is, in fact, smaller than that by the traditional t-test, illustrating the merit of our method.

Learning Temporal Alignment Uncertainty for Efficient Event Detection

In this paper we tackle the problem of efficient video event detection. We argue that linear detection functions should be preferred in this regard due to their scalability and efficiency during estimation and evaluation. A popular approach in this regard is to represent a sequence using a bag of words (BOW) representation due to its: (i) fixed dimensionality irrespective of the sequence length, and (ii) its ability to compactly model the statistics in the sequence. A drawback to the BOW representation, however, is the intrinsic destruction of the temporal ordering information. In this paper we propose a new representation that leverages the uncertainty in relative temporal alignments between pairs of sequences while not destroying temporal ordering. Our representation, like BOW, is of a fixed dimensionality making it easily integrated with a linear detection function. Extensive experiments on CK+, 6DMG, and UvA-NEMO databases show significant performance improvements across both isolated and continuous event detection tasks.

Rapid isolation and detection of erythropoietin in blood plasma by magnetic core gold nanoparticles and portable Raman spectroscopy

Isolating, purifying, and identifying proteins in complex biological matrices is often difficult, time consuming, and unreliable. Herein we describe a rapid screening technique for proteins in biological matrices that combines selective protein isolation with direct surface enhanced Raman spectroscopy (SERS) detection. Magnetic core gold nanoparticles were synthesised, characterised, and subsequently functionalized with recombinant human erythropoietin (rHuEPO)-specific antibody. The functionalized nanoparticles were used to capture rHuEPO from horse blood plasma within 15 minutes. The selective binding between the protein and the functionalized nanoparticles was monitored by SERS. The purified protein was then released from the nanoparticles’ surface and directly spectroscopically identified on a commercial nanopillar SERS substrate. ELISA independently confirmed the SERS identification and quantified the released rHuEPO. Finally, the direct SERS detection of the extracted protein was successfully demonstrated for in-field screening by a handheld Raman spectrometer within 1 minute sample measurement time.

Teledermatology: Its use in the detection and management of actinic keratosis

Teledermatology can profoundly improve access to medical services for those who may have limited access to dermatology due to workforce shortages, distance to providers, or limitations in their mobility. Two common ways of teledermatology are differentiated: life synchronous, where patient and doctor communicate directly, or store and forward asynchronous methods, where the patient and doctor provide and assess the medical information independently. Teledermatology has been tested for its safety, feasibility and accuracy for a number of dermatological conditions, including the early detection of skin cancer and is usually safe, feasible and accurate. Studies reported somewhat better results for synchronous than asynchronous methods, possibly because of loss of information if no direct patient doctor contact is feasible. However asynchronous methods are easier to organize, require less sophisticated technology and are more widely accessible, and are more convenient for both patients and doctors. No study to date focused solely on teledermatology of actinic keratosis, but such lesions are typically found during teledermatology examinations for other main target lesions. In studies where such results were reported, actinic keratoses seemed to be readily identifiable for teledermatologists and adequate management and treatment can be suggested within remote consultations.

Effect of position on precision within the recording field of a markerless motion capture system

Introduction Markerless motion capture systems are relatively new devices that can significantly speed up capturing full body motion. A precision of the assessment of the finger’s position with this type of equipment was evaluated at 17.30 ± 9.56 mm when compare to an active marker system [1]. The Microsoft Kinect was proposed to standardized and enhanced clinical evaluation of patients with hemiplegic cerebral palsy [2]. Markerless motion capture systems have the potential to be used in a clinical setting for movement analysis, as well as for large cohort research. However, the precision of such system needs to be characterized. Global objectives • To assess the precision within the recording field of the markerless motion capture system Openstage 2 (Organic Motion, NY). • To compare the markerless motion capture system with an optoelectric motion capture system with active markers. Specific objectives • To assess the noise of a static body at 13 different location within the recording field of the markerless motion capture system. • To assess the smallest oscillation detected by the markerless motion capture system. • To assess the difference between both systems regarding the body joint angle measurement. Methods Equipment • OpenStage® 2 (Organic Motion, NY) o Markerless motion capture system o 16 video cameras (acquisition rate : 60Hz) o Recording zone : 4m * 5m * 2.4m (depth * width * height) o Provide position and angle of 23 different body segments • VisualeyezTM VZ4000 (PhoeniX Technologies Incorporated, BC) o Optoelectric motion capture system with active markers o 4 trackers system (total of 12 cameras) o Accuracy : 0.5~0.7mm Protocol & Analysis • Static noise: o Motion recording of an humanoid mannequin was done in 13 different locations o RMSE was calculated for each segment in each location • Smallest oscillation detected: o Small oscillations were induced to the humanoid mannequin and motion was recorded until it stopped. o Correlation between the displacement of the head recorded by both systems was measured. A corresponding magnitude was also measured. • Body joints angle: o Body motion was recorded simultaneously with both systems (left side only). o 6 participants (3 females; 32.7 ± 9.4 years old) • Tasks: Walk, Squat, Shoulder flexion & abduction, Elbow flexion, Wrist extension, Pronation / supination (not in results), Head flexion & rotation (not in results), Leg rotation (not in results), Trunk rotation (not in results) o Several body joint angles were measured with both systems. o RMSE was calculated between signals of both systems. Results Conclusion Results show that the Organic Motion markerless system has the potential to be used for assessment of clinical motor symptoms or motor performances However, the following points should be considered: • Precision of the Openstage system varied within the recording field. • Precision is not constant between limb segments. • The error seems to be higher close to the range of motion extremities.

A coupled SPH-DEM approach to model the interactions between multiple red blood cells in motion in capillaries

Red blood cells (RBCs) are the most common type of blood cells in the blood and 99% of the blood cells are RBCs. During the circulation of blood in the cardiovascular network, RBCs squeeze through the tiny blood vessels (capillaries). They exhibit various types of motions and deformed shapes, when flowing through these capillaries with diameters varying between 5 10 µm. RBCs occupy about 45 % of the whole blood volume and the interaction between the RBCs directly influences on the motion and the deformation of the RBCs. However, most of the previous numerical studies have explored the motion and deformation of a single RBC when the interaction between RBCs has been neglected. In this study, motion and deformation of two 2D (two-dimensional) RBCs in capillaries are comprehensively explored using a coupled smoothed particle hydrodynamics (SPH) and discrete element method (DEM) model. In order to clearly model the interactions between RBCs, only two RBCs are considered in this study even though blood with RBCs is continuously flowing through the blood vessels. A spring network based on the DEM is employed to model the viscoelastic membrane of the RBC while the inside and outside fluid of RBC is modelled by SPH. The effect of the initial distance between two RBCs, membrane bending stiffness (Kb) of one RBC and undeformed diameter of one RBC on the motion and deformation of both RBCs in a uniform capillary is studied. Finally, the deformation behavior of two RBCs in a stenosed capillary is also examined. Simulation results reveal that the interaction between RBCs has significant influence on their motion and deformation.