Search

One of the major drawbacks to the large scale application of magnetic support based separations in biotechnology is the astronomic price of large quantities of commercially available magnetic support materials. The aim of this work has been to develop a simple, fast and scalable ‘one-pot’ manufacturing route to cheap magnetic adsorbents to be used in High-Gradient Magnetic Fishing, Magnetically Enhanced Press Filtration and Magnetically Enhanced Centrifugation systems. An inverse liquid-liquid two phase polymerization technique employing low cost chemicals combined with the use of a rotor stator type high shear mixer was systematically investigated. When the ‘ideal’ manufacturing conditions were identified, nonporous superparamagnetic composite supports (3 .m average size; M\(_s\): 34 A m\(^2\) kg\(^{-1}\); magnetic content: 44.3%) were successfully created. Following further optimization, the ‘one-pot’ manufacturing procedure was used to produce anion exchange magnetic adsorbents, and their binding capacity was determined using human serum albumin as a model species (Q\(_{max}\)=46.1 mg of protein per gram of adsorbent; K\(_d\)=2.52 .M). Further attempts to create cation exchange magnetic supports were unsuccessful mainly due to chemical incompatibility issues between the reaction chemicals and the ferrofluid used.

Autonomous vehicle (AV) prototypes have been deployed in increasingly varied environments in recent years. An AV must be able to reliably detect and predict the future motion of traffic participants to maintain safe operation based on data collected from high-quality onboard sensors. Sensors such as camera and LiDAR generate high-bandwidth data that requires substantial computational and memory resources. To address these AV challenges, this thesis investigates three related problems: 1) What will the observed traffic participants do? 2) Is an anomalous traffic event likely to happen in near future? and 3) How should we collect fleet-wide high-bandwidth data based on 1) and 2) over the long-term? The first problem is addressed with future traffic trajectory and pedestrian behavior prediction. We propose a future object localization (FOL) method for trajectory prediction in first person videos (FPV). FOL encodes heterogeneous observations including bounding boxes, optical flow features and ego camera motions with multi-stream recurrent neural networks (RNN) to predict future trajectories. Because FOL does not consider multi-modal future trajectories, its accuracy suffers from accumulated RNN prediction error. We then introduce BiTraP, a goal-conditioned bidirectional multi-modal trajectory prediction method. BiTraP estimates multi-modal trajectories and uses a novel bi-directional decoder and loss to improve longer-term trajectory prediction accuracy. We show that different choices of non-parametric versus parametric target models directly influence predicted multi-modal trajectory distributions. Experiments with two FPV and six bird's-eye view (BEV) datasets show the effectiveness of our methods compared to state-of-the-art. We define pedestrian behavior prediction as a combination of action and intent. We hypothesize that current and future actions are strong intent priors and propose a multi-task learning RNN encoder-decoder network to detect and predict future pedestrian actions and street crossing intent. Experimental results show that one task helps the other so they together achieve state-of-the-art performance on published datasets. To identify likely traffic anomaly events, we introduce an unsupervised video anomaly detection (VAD) method based on trajectories. We predict locations of traffic participants over a near-term future horizon and monitor accuracy and consistency of these predictions as evidence of an anomaly. Inconsistent predictions tend to indicate an anomaly has happened or is about to occur. A supervised video action recognition method can then be applied to classify detected anomalies. We introduce a spatial-temporal area under curve (STAUC) metric as a supplement to the existing area under curve (AUC) evaluation and show it captures how well a model detects temporal and spatial locations of anomalous events. Experimental results show the proposed method and consistency-based anomaly score are more robust to moving cameras than image generation based methods; our method achieves state-of-the-art performance over AUC and STAUC metrics. VAD and action recognition support event-of-interest (EOI) distinction from normal driving data. We introduce a Smart Black Box (SBB), an intelligent event data recorder, to prioritize EOI data in long-term driving. The SBB compresses high-bandwidth data based on EOI potential and on-board storage limits. The SBB is designed to prioritize newer and anomalous driving data and discard older and normal data. An optimal compression factor is selected based on the trade-off between data value and storage cost. Experiments in a traffic simulator and with real-world datasets show the efficiency and effectiveness of using a SBB to collect high-quality videos over long-term driving.

Ethnic media is said to perform a number of roles, including: 1) protecting and (re) constructing ethnic culture and identity, 2) providing knowledge and information to newcomers and 3) giving an independent space which allows ethnic communities to represent themselves in their own words. In New Zealand, the number of Chinese migrants has steadily increased since the introduction of the 1987 Immigration Act, but the portrayal of the Chinese diaspora in its dominant media has often been invisible and negative. This is difficult for Chinese people to tolerate because they are afraid to ‘lose face’, which is an important part of traditional Chinese social ethics and refers to the fear of public shame. Along with other Chinese cultural characteristics of collectivism, members of the Chinese diaspora may consider that bad descriptions would bind them together and affect other people’s impression about their community. Under the circumstance, Chinese ethnic media may play an important role in assuaging people’s feelings and counteracting the negative influences within the Chinese community. This study employed two qualitative research methods, content analysis and semi-structured interviews, to probe the way in which one Chinese ethnic online media, SkyKiwi.com, responded to the representation of the Chinese people in the New Zealand Herald[the Herald].The research is based on two case studies. The first case involves Auckland Mayor Len Brown’s affair with Bevan Chuang in 2013. The second case is based on the reporting on an incident during which a group identified by the Herald as Chinese tourists ate at a charity Christmas lunch in 2012. A comparison of reports about these events in both media outlets, indicates a tendency of the Herald to focus on more dramatic descriptions of celebrities and to be more connected with the dominant, Pākehā group. The findings showed SkyKiwi.com emphasises vicarious experience, which is the involvement of the audience to engage in a news story by highlighting the emotional elements, in its news reports and the way it used online sources. Two strategies, including the selection of information for pre-determined goals and the audience-oriented description, can be identified in these comparisons. These strategies explain the important role SkyKiwi.com played in response to the portrayals of Chinese people in the Herald, including guiding the audience to understand both events from another angle and providing an independent, alternative media space in which the Chinese diaspora in New Zealand could represent themselves in their own words.