Search

In this paper we address the basic limitation of SiamMask - the state of the art single object tracking and segmentation algorithm. SiamMask requires semi-supervision in that it needs a bounding box to be drawn manually around the object that has to be tracked. This is however not always possible or feasible, and slows down the pipeline even in the best case. We overcome this limitation by using state-of-the-art object detection algorithms: Detectron2 and YOLO to automatically detect the object and then track using SiamMask. The main purpose of this study is to devise an efficient technique for an end-to-end object detection and tracking, which can then be used in other applications like self-driving cars, etc. We compared different approaches using current state-of-the-art tools for time and detection efficiency. One of the secondary aim was to test how the two approaches perform on different types of datasets. We note that YOLO gives better and more meaningful detection of objects in the scene. However, Detectron2 gives a higher detection speed than YOLO, making the overall detection and tracking process faster.

The dynamic network topology of Flying Ad hoc Networks (FANETs) leads to challenges during communications. This becomes complicated when dealing with multiple transmission paths and relays. Conventional routing protocols proposed fail to address the dynamic issues inherent in FANETs. This work addresses it by diversifying the selection of relay in order to establish the significance of cooperative diversity technique. Inspiration is taken from bio inspired computing which assists in finding solutions to several challenging tasks. The natural behavior of the different species leads to elevated design concepts of protocols. This paper proposes GW-COOP (Gray Wolf Algorithm using Cooperative Diversity Technique) routing protocol for FANETs. This protocol consists of Gray Wolf Optimizer (GWO) that drives gray wolves’ social hierarchy and collaboration technique. First, we opt the design implementation of gray wolves natural posture GWO to handle flying node requirements. Second, we envision the idea of cooperative diversity using two relays to sustain either of the source-to-destination links. The design of protocol is novel in that previously such cooperation techniques are not employed in FANETs. Moreover, the concept of two relays with bio inspired algorithm is first time proposed here. In order to establish the performance of the GW-COOP protocol, this paper compares GW-COOP with two different protocols BAT-COOP (Bat Algorithm using Cooperative Diversity Technique) and BAT-FANETs protocols. The simulation results show that the GW-COOP protocol outperforms BAT-COOP protocol in terms of transmission loss, energy consumption, link delay and packet loss ratio. The results indicate improved GW-COOP performance over BAT-FANET and BAT-COOP. Approximately 67% and 52% reduction is observed in transmission losses compared to BAT-FANET and BAT-COOP respectively. Moreover, a decrease of 61% and 54% in terms of energy consumption, 24% and 9% in terms of link delay, and 58% and 48% reduction in terms of packet loss ratio, is also observed.

Evolution of drug-resistant microbial species is one of the major challenges to global health. Development of new antimicrobial treatments such as antimicrobial peptides needs to be accelerated to combat this threat. However, the discovery of novel antimicrobial peptides is hampered by low-throughput biochemical assays. Computational techniques can be used for rapid screening of promising antimicrobial peptide candidates prior to testing in the wet lab. The vast majority of existing antimicrobial peptide predictors are non-targeted in nature, i.e., they can predict whether a given peptide sequence is antimicrobial, but they are unable to predict whether the sequence can target a particular microbial species. In this work, we have used zero and few shot machine learning to develop a targeted antimicrobial peptide activity predictor called AMP0. The proposed predictor takes the sequence of a peptide and any N/C-termini modifications together with the genomic sequence of a microbial species to generate targeted predictions. Cross-validation results show that the proposed scheme is particularly effective for targeted antimicrobial prediction in comparison to existing approaches and can be used for screening potential antimicrobial peptides in a targeted manner with only a small number of training examples for novel species. AMP0 webserver is available at http://ampzero.pythonanywhere.com.

Neurostimulation with multiple scalp electrodes has shown enhanced effects in recent studies. However, visualizations of stimulation-induced internal current distributions in brain is only possible through simulated current distributions obtained from computer model of human head. While magnetic resonance current density imaging (MRCDI) has a potential for direct in-vivo measurement of currents induced in brain with multi-electrode stimulation, existing MRCDI methods are only developed for two-electrode neurostimulation. A major bottleneck is the lack of a current switching device which is typically used to convert the DC current of neurostimulation devices into user-defined waveforms of positive and negative polarity with delays between them. In this work, we present a design of a four-electrode current switching device to enable simultaneous switching of current flowing through multiple scalp electrodes.

Functional connectivity (FC) mapping from resting-state functional magnetic resonance imaging (rsfMRI) data is a widely used technique to characterize the brain abnormalities in mental health disorders. Using atlases for brain parcellation is an important intermediate step in calculation of FC maps. Atlases with varying resolution (number of nodes in an atlas) have been deployed by researchers to study the abnormal brain functions in Schizophrenia. In this work, we compared the variations in FC maps of Schizophrenic brains obtained from three different atlases: AAL atlas (2002), Dosenbach atlas (2010), and the Brainnetome atlas (2016). To evaluate the atlas-dependent variations in FC maps, we relied on the capability of the features of FC maps in accurately classifying a given data into healthy or Schizophrenia group. Our results indicate that the high-resolution Dosenbach and Brainnetome atlases perform better than AAL atlas in terms of the accuracy, sensitivity and specificity of the SVM classifier.

Machine vision tasks present challenges for resource constrained edge devices, particularly as they execute multiple tasks with variable workloads. A robust approach that can dynamically adapt in runtime while maintaining the maximum quality of service (QoS) within resource constraints, is needed. The paper presents a lightweight approach that monitors the runtime workload constraint and leverages accuracy-throughput trade-off. Optimisation techniques are included which find the configurations for each task for optimal accuracy, energy and memory and manages transparent switching between configurations. For an accuracy drop of 1%, we show a 1.6x higher achieved frame processing rate with further improvements possible at lower accuracy.

The development of deep segmentation models for computational pathology (CPath) can help foster the investigation of interpretable morphological biomarkers. Yet, there is a major bottleneck in the success of such approaches because supervised deep learning models require an abundance of accurately labelled data. This issue is exacerbated in the field of CPath because the generation of detailed annotations usually demands the input of a pathologist to be able to distinguish between different tissue constructs and nuclei. Manually labelling nuclei may not be a feasible approach for collecting large-scale annotated datasets, especially when a single image region can contain thousands of different cells. However, solely relying on automatic generation of annotations will limit the accuracy and reliability of ground truth. Therefore, to help overcome the above challenges, we propose a multi-stage annotation pipeline to enable the collection of large-scale datasets for histology image analysis, with pathologist-in-the-loop refinement steps. Using this pipeline, we generate the largest known nuclear instance segmentation and classification dataset, containing nearly half a million labelled nuclei in H&E stained colon tissue. We have released the dataset and encourage the research community to utilise it to drive forward the development of downstream cell-based models in CPath.

Cellular composition prediction, i.e., predicting the presence and counts of different types of cells in the tumor microenvironment from a digitized image of a Hematoxylin and Eosin (H&E) stained tissue section can be used for various tasks in computational pathology such as the analysis of cellular topology and interactions, subtype prediction, survival analysis, etc. In this work, we propose an image-based cellular composition predictor (ALBRT) which can accurately predict the presence and counts of different types of cells in a given image patch. ALBRT, by its contrastive-learning inspired design, learns a compact and rotation-invariant feature representation that is then used for cellular composition prediction of different cell types. It offers significant improvement over existing state-of-the-art approaches for cell classification and counting. The patch-level feature representation learned by ALBRT is transferrable for cellular composition analysis over novel datasets and can also be utilized for downstream prediction tasks in CPath as well. The code and the inference webserver for the proposed method are available at the URL: https://github.com/engrodawood/ALBRT., 11 pages, 5 figures

Skin segmentation plays an important role in human activity recognition, video surveillance, hand gesture identification, face detection, human tracking and robotic surgery. The accurate segmentation of the skin is necessary to recognize the human activity. Segmentation of skin is easy to realize in ideal situations because of similar backgrounds. But it becomes complicated because of presence of skin-like pixels, background illuminations, and certain changes in environment. These problems are addressed by incorporating preprocessing stages in current studies, but this raises the total cost of the system. However, there are some limitations associated with these methods in terms of accuracy and processing speed. In this work, we propose a skin semantic segmentation network (SSS-Net) that is able to capture the multi-scale contextual information and refines the segmentation results especially along object boundaries. Moreover our network helps to reduce the cost of the preprocessing as well. We have performed experiments on the five open datasets of human activity recognition for the segmentation of skin. Experimental results show SSS-Net outperforms the state-of-the-art methods in skin segmentation in terms of accuracies.

Brick kiln areas being home to practices of forced labor and environmental pollution are difficult to detect. This paper explores the automatic detection and localization of brick kilns in satellite images by the implementation of Deep Learning Techniques to aid the government to take action against forced labour and environmental pollution. An artificially intelligent model was trained to automatically detect and create a bounding box around the kilns in satellite images from the South Asian brick belt. The experimentation started with the use of basic convolutional neural networks to identify kilns in an image and then progressed to the more advanced You Only Look Once (YOLOv3) algorithm to draw bounding boxes across the kilns. Our research enabled us to successfully build a convolutional neural network for the identification of kilns which returned an accuracy of 97.27%. For the localization of kilns using the bounding box approach, our research enabled us to implement the YOLOv3 algorithm which returned an average loss of 0.13 and an average IOU of 0.75. The work done in this project has been promising thus far and can be further enhanced to build a Graphical User Interface to help governments to take action against the social issues of bonded labor and environmental issues that are prevalent within brick kilns.

This paper aims to initiate a discussion around benchmarking data management systems with machine-learned components. Traditional benchmarks such as TPC or YCSB are insufficient to analyze and understand these learned systems because they evaluate the performance under a stable workload and data distribution. Learned systems automatically specialize and adapt database components to a changing workload, database, and execution environment, thereby making conventional metrics such as average throughput ill-suited to understand their performance fully. Moreover, the standard cost-per-performance metrics fail to account for essential trade-offs related to the training cost of models and the elimination of manual database tuning. We present several ideas for designing new benchmarks that are better suited to evaluate learned systems. The main challenges entail developing new metrics to capture the particularities of learned systems and ensuring that benchmark results remain comparable across many deployments with wide-ranging designs.