Jinglan Zhang

Bio: Jinglan Zhang is an academic researcher from Queensland University of Technology. The author has contributed to research in topics: Deep learning & Feature extraction. The author has an hindex of 22, co-authored 168 publications receiving 1957 citations.

Review of deep learning: concepts, CNN architectures, challenges, applications, future directions

Abstract: In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

Towards automatic power line detection for a UAV surveillance system using pulse coupled neural filter and an improved Hough transform

Abstract: Spatial information captured from optical remote sensors on board unmanned aerial vehicles (UAVs) has great potential in automatic surveillance of electrical infrastructure. For an automatic vision-based power line inspection system, detecting power lines from a cluttered background is one of the most important and challenging tasks. In this paper, a novel method is proposed, specifically for power line detection from aerial images. A pulse coupled neural filter is developed to remove background noise and generate an edge map prior to the Hough transform being employed to detect straight lines. An improved Hough transform is used by performing knowledge-based line clustering in Hough space to refine the detection results. The experiment on real image data captured from a UAV platform demonstrates that the proposed approach is effective for automatic power line detection.

Towards a Better Understanding of Transfer Learning for Medical Imaging: A Case Study

Abstract: One of the main challenges of employing deep learning models in the field of medicine is a lack of training data due to difficulty in collecting and labeling data, which needs to be performed by experts. To overcome this drawback, transfer learning (TL) has been utilized to solve several medical imaging tasks using pre-trained state-of-the-art models from the ImageNet dataset. However, there are primary divergences in data features, sizes, and task characteristics between the natural image classification and the targeted medical imaging tasks. Therefore, TL can slightly improve performance if the source domain is completely different from the target domain. In this paper, we explore the benefit of TL from the same and different domains of the target tasks. To do so, we designed a deep convolutional neural network (DCNN) model that integrates three ideas including traditional and parallel convolutional layers and residual connections along with global average pooling. We trained the proposed model against several scenarios. We utilized the same and different domain TL with the diabetic foot ulcer (DFU) classification task and with the animal classification task. We have empirically shown that the source of TL from the same domain can significantly improve the performance considering a reduced number of images in the same domain of the target dataset. The proposed model with the DFU dataset achieved F1-score value of 86.6% when trained from scratch, 89.4% with TL from a different domain of the targeted dataset, and 97.6% with TL from the same domain of the targeted dataset.

Sensor Network for the Monitoring of Ecosystem: Bird Species Recognition

Abstract: In this paper, we investigated the performance of bird species recognition using neural networks with different preprocessing methods and different sets of features. Context neural network architecture was designed to embed the dynamic nature of bird songs into inputs. We devised a noise reduction algorithm and effectively applied it to enhance bird species recognition. The performance of the context neural network architecture was comparatively evaluated with linear/mel frequency cepstral coefficients and promising experimental results were achieved.

Knowledge-based power line detection for UAV surveillance and inspection systems

Abstract: Spatial information captured from optical remote sensors on board unmanned aerial vehicles (UAVs) has great potential in the automatic surveillance of electrical power infrastructure. For an automatic vision based power line inspection system, detecting power lines from cluttered background an important and challenging task. In this paper, we propose a knowledge-based power line detection method for a vision based UAV surveillance and inspection system. A PCNN filter is developed to remove background noise from the images prior to the Hough transform being employed to detect straight lines. Finally knowledge based line clustering is applied to refine the detection results. The experiment on real image data captured from a UAV platform demonstrates that the proposed approach is effective.

Pattern Recognition and Machine Learning

Abstract: Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Design Prototypes: A Knowledge Representation Schema for Design

Abstract: A prevalent and pervasive view of designing is that it can be modeled using variables and decisions made about what values should be taken by these variables. The activity of designing is carried out with the expectation that the designed artifact will operate in the natural world and the social world. These worlds impose constraints on the variables and their values; so, design could be described as a goal-oriented, constrained, decision- making activity. However, design distinguish- es itself from other similarly described activities not only by its domain but also by additional necessary features. Designing involves exploration, exploring what variables might be appropriate. The process of explo- ration involves both goal variables and deci- sion variables. In addition, designing involves learning: Part of the exploration activity is learning about emerging features as a design proceeds. Finally, design activity occurs within two contexts: the context within which the designer operates and the context produced by the developing design itself. The designer’s perception of what the context is affects the implication of the context on the design. The context shifts as the designer’s perceptions change. Design activity can be now characterized as a goal-oriented, con- strained, decision-making, exploration, and learning activity that operates within a con- text that depends on the designer’s percep- tion of the context.

Salient Object Detection: A Benchmark

Abstract: We extensively compare, qualitatively and quantitatively, 41 state-of-the-art models (29 salient object detection, 10 fixation prediction, 1 objectness, and 1 baseline) over seven challenging data sets for the purpose of benchmarking salient object detection and segmentation methods. From the results obtained so far, our evaluation shows a consistent rapid progress over the last few years in terms of both accuracy and running time. The top contenders in this benchmark significantly outperform the models identified as the best in the previous benchmark conducted three years ago. We find that the models designed specifically for salient object detection generally work better than models in closely related areas, which in turn provides a precise definition and suggests an appropriate treatment of this problem that distinguishes it from other problems. In particular, we analyze the influences of center bias and scene complexity in model performance, which, along with the hard cases for the state-of-the-art models, provide useful hints toward constructing more challenging large-scale data sets and better saliency models. Finally, we propose probable solutions for tackling several open problems, such as evaluation scores and data set bias, which also suggest future research directions in the rapidly growing field of salient object detection.