No description available.

Flame and smoke image processing and analysis could improve performance to detect smoke or fire and identify many complicated fire hazards, eventually to help firefighters to fight fires safely. Deep Learning applied to image processing has been prevailing in recent years among image-related research fields. Fire safety researchers also brought it into their studies due to its leading performance in image-related tasks and statistical analysis. From the perspective of input data type, traditional fire research is based on simple mathematical regressions or empirical correlations relying on sensor data, such as temperature. However, data from advanced vision devices or sensors can be analyzed by applying deep learning beyond auxiliary methods in data processing and analysis. Deep Learning has a bigger capacity in non-linear problems, especially in high-dimensional spaces, such as flame and smoke image processing. We propose a video-based real-time smoke and flame analysis system with deep learning networks and fire safety knowledge. It takes videos of fire as input and produces analysis and prediction for flashover of fire. Our system consists of four modules. The Color2IR Conversion module is made by deep neural networks to convert RGB video frames into InfraRed (IR) frames, which could provide important thermal information of fire. Thermal information is critically important for fire hazard detection. For example, 600 °C marks the start of a flashover. As RGB cameras cannot capture thermal information, we propose an image conversion module from RGB to IR images. The core of this conversion is a new network that we innovatively proposed: Dual-Attention Generative Adversarial Network (DAGAN), and it is trained using a pair of RGB and IR images. Next, Video Semantic Segmentation Module helps extract flame and smoke areas from the scene in the RGB video frames. We innovated to use synthetic RGB video data generated and captured from 3D modeling software for data augmentation. After that, a Video Prediction Module takes the RGB video frames and IR frames as input and produces predictions of the subsequent frames of their scenes. Finally, a Fire Knowledge Analysis Module predicts if flashover is coming or not, based on fire knowledge criteria such as thermal information extracted from IR images, temperature increase rate, the flashover occurrence temperature, and increase rate of lowest temperature. For our contributions and innovations, we introduce a novel network, DAGAN, by applying foreground and background attention mechanisms in the image conversion module to help reduce the hardware device requirement for flashover prediction. Besides, we also make use of combination of thermal information from IR images and segmentation information from RGB images in our system for flame and smoke analysis. We also apply a hybrid design of deep neural networks and a knowledge-based system to achieve high accuracy. Moreover, data augmentation is also applied on the Video Semantic Segmentation Module by introducing synthetic video data for training. The test results of flashover prediction show that our system has leading places quantitative and qualitative in terms of various metrics compared with other existing approaches. It can give a flashover prediction as early as 51 seconds with 94.5% accuracy before it happens.

Deep learning models are dominating performance across a wide variety of tasks. From protein folding to computer vision to voice recognition, deep learning is changing the way we interact with data. The field of natural products, and more specifically genomic mining, has been slow to adapt to these new technological innovations. As we are in the midst of a data explosion, it is not for lack of training data. Instead, it is due to the lack of a blueprint demonstrating how to correctly integrate these models to maximise performance and inference. During my PhD, I showcase the use of large language models across a variety of data domains to improve common workflows in the field of natural product drug discovery. I improved natural product scaffold comparison by representing molecules as sentences. I developed a series of deep learning models to replace archaic technologies and create a more scalable genomic mining pipeline decreasing running times by 8X. I integrated deep learning-based genomic and enzymatic inference into legacy tooling to improve the quality of short-read assemblies. I also demonstrate how intelligent querying of multi-omic datasets can be used to facilitate the gene signature prediction of encoded microbial metabolites. The models and workflows I developed are wide in scope with the hopes of blueprinting how these industry standard tools can be applied across the entirety of natural product drug discovery. / Thesis / Doctor of Philosophy (PhD)

Sponges play an important role in many marine habitats and are crucial for maintaining the deep-sea marine ecosystems. However, there is a knowledge gap in the field of sponge biodiversity due to insufficient exploration of the deep-sea, and the probability of finding new species is fairly high. The deep waters of the Indian and Pacific Oceans are particularly poorly explored, and their sponge diversity is barely known. During the KANADEEP 2 expedition in the south of New Caledonia a large collection of Demospongiae and Hexactinellida was gathered to assess the sponge biodiversity. In this project 110 Demosponges were investigated morphologically and identified to the order level. Twenty-one of these were selected for further identification and description in terms of shape and geometry of spicules using light microscopy. The DNA was extracted from eight specimens for the confirmation of new species. Amplification of CO1 Folmer fragment was conducted using PCR. The resulting PCR-products were analyzed using gel electrophoresis and DNA-sequencing. The twenty-one specimens were assigned to the genera Tethya, Stupenda and Geodia belonging to the order Tetractinellida, which was found to be the dominant order in the deep waters in the New Caledonia region. There are potentially six new species amongst the studied specimens. However, it has to be confirmed with molecular analysis of specific markers. The morphological analysis of 21 specimens collected during the KANADEEP 2 expedition demonstrated that 17 specimens belonged to Geodia, two specimens to Tethya, and two specimens to Stupenda.

No description available.

No description available.

Gait recognition systems have gained tremendous attention due to its potential applications in healthcare, criminal investigation, sports biomechanics, and so forth. A new solution to gait recognition tasks can be provided by wearable sensors integrated in wearable objects or mobile devices. In this research a sock prototype designed with embedded soft robotic sensors (SRS) is implemented to measure foot ankle kinematic and kinetic data during three experiments designed to track participants’ feet ankle movement. Deep learning and statistical methods have been employed to model SRS data against Motion capture system (MoCap) to determine their ability to provide accurate kinematic and kinetic data using SRS measurements. In the first study, the capacitance of SRS related to foot-ankle basic movements was quantified during the gait movements of twenty participants on a flat surface and a cross-sloped surface. I have conducted another study regarding kinematic features in which deep learning models were trained to estimate the joint angles in sagittal and frontal planes measured by a MoCap system. Participant-specific models were established for ten healthy subjects walking on a treadmill. The prototype was tested at various walking speeds to assess its ability to track movements for multiple speeds and generalize models for estimating joint angles in sagittal and frontal planes. The focus of the last study is measuring the kinetic features and the goal is determining the validity of SRS measurements, to this end the pressure data measured with SRS embedded into the sock prototype would be compared with the force plate data.

Software requirements classification, the practice of categorizing requirements by their type or purpose, can improve organization and transparency in the requirements engineering process and thus promote requirement fulfillment and software project completion. Requirements classification automation is a prominent area of research as automation can alleviate the tediousness of manual labeling and loosen its necessity for domain-expertise. This thesis explores the application of deep learning techniques on software requirements classification, specifically the use of word embeddings for document representation when training a convolutional neural network (CNN). As past research endeavors mainly utilize information retrieval and traditional machine learning techniques, we entertain the potential of deep learning on this particular task. With the support of learning libraries such as TensorFlow and Scikit-Learn and word embedding models such as word2vec and fastText, we build a Python system that trains and validates configurations of Naïve Bayes and CNN requirements classifiers. Applying our system to a suite of experiments on two well-studied requirements datasets, we recreate or establish the Naïve Bayes baselines and evaluate the impact of CNNs equipped with word embeddings trained from scratch versus word embeddings pre-trained on Big Data.

<p>Roller bearings are the common components used in the mechanical systems for mechanical processing and production. The running state of roller bearings often determines the machining accuracy and productivity on a manufacturing line. Roller bearing failure may lead to the shutdown of production lines, resulting in serious economic losses. Therefore, the research on roller bearing fault diagnosis has a great value. This thesis research first proposes a method of signal frequency spectral resampling to tackle the problem of bearing fault detection at different rotating speeds using a single speed dataset for training the network such as the one dimensional convolutional neural network (1D CNN). Second, this research work proposes a technique to connect the graph structures constructed from spectral components of the different bearing fault frequency bands into a sparse graph structure, so that the fault identification can be carried out effectively through a graph neural network in terms of the computation load and classification rate. Finally, the frequency spectral resampling method for feature extraction is validated using our self-collected datasets. The performance of the graph neural network with our proposed sparse graph structure is validated using the Case Western Reserve University (CWRU) dataset as well as our self-collected datasets. The results show that our proposed method achieves higher bearing fault classification accuracy than those recently proposed by other researchers using machine learning approaches and neural networks.</p>

Tractography is widely used in the brain connectivity study from diffusion magnetic resonance imaging data. However, lack of ground truth and plenty of anatomically implausible streamlines in the tractograms have caused challenges and concerns in the use of tractograms such as brain connectivity study. Tractogram filtering methods have been developed to remove the faulty connections. In this study, we focus on one of these filtering methods, Convex Optimization Modeling for Microstructure Informed Tractography (COMMIT), which tries to find a set of streamlines that best reconstruct the diffusion magnetic resonance imaging data with global optimization approach. There are biases with this method when assessing individual streamlines. So a method named randomized COMMIT(rCOMMIT) is proposed to obtain multiple assessments for each streamline. The acceptance rate from this method is introduced to the streamlines and divides them into three groups, which are regarded as pseudo ground truth from rCOMMIT. Therefore, the neural networks are able to train on the pseudo ground truth on classification tasks. The trained classifiers distinguish the obtained groups of plausible and implausible streamlines with accuracy around 77%. Following the same methodology, the results from rCOMMIT and randomized SIFT are compared. The intersections between two methods are analyzed with neural networks as well, which achieve accuracy around 87% in binary task between plausible and implausible streamlines.