Vijayan K. Asari

R. Wes Baldwin, Vijayan Asari, Keigo Hirakawa

We present a novel Fourier camera, an in-hardware optical compression of high-speed frames employing pixel-level sign-coded exposure where pixel intensities temporally modulated as positive and negative exposure are combined to yield Hadamard coefficients. The orthogonality of Walsh functions ensures that the noise is not amplified during high-speed frame reconstruction, making it a much more attractive option for coded exposure systems aimed at very high frame rate operation. Frame reconstruction is carried out by a single-pass demosaicking of the spatially multiplexed Walsh functions in a lattice arrangement, significantly reducing the computational complexity. The simulation prototype confirms the improved robustness to noise compared to the binary-coded exposure patterns, such as one-hot encoding and pseudo-random encoding. Our hardware prototype demonstrated the reconstruction of 4kHz frames of a moving scene lit by ambient light only.

Dhaval D. Kadia, Md Zahangir Alom, Ranga Burada et al.

3D lung segmentation is essential since it processes the volumetric information of the lungs, removes the unnecessary areas of the scan, and segments the actual area of the lungs in a 3D volume. Recently, the deep learning model, such as U-Net outperforms other network architectures for biomedical image segmentation. In this paper, we propose a novel model, namely, Recurrent Residual 3D U-Net (R2U3D), for the 3D lung segmentation task. In particular, the proposed model integrates 3D convolution into the Recurrent Residual Neural Network based on U-Net. It helps learn spatial dependencies in 3D and increases the propagation of 3D volumetric information. The proposed R2U3D network is trained on the publicly available dataset LUNA16 and it achieves state-of-the-art performance on both LUNA16 (testing set) and VESSEL12 dataset. In addition, we show that training the R2U3D model with a smaller number of CT scans, i.e., 100 scans, without applying data augmentation achieves an outstanding result in terms of Soft Dice Similarity Coefficient (Soft-DSC) of 0.9920.

R. Wes Baldwin, Ruixu Liu, Mohammed Almatrafi et al.

Event cameras are an exciting, new sensor modality enabling high-speed imaging with extremely low-latency and wide dynamic range. Unfortunately, most machine learning architectures are not designed to directly handle sparse data, like that generated from event cameras. Many state-of-the-art algorithms for event cameras rely on interpolated event representations - obscuring crucial timing information, increasing the data volume, and limiting overall network performance. This paper details an event representation called Time-Ordered Recent Event (TORE) volumes. TORE volumes are designed to compactly store raw spike timing information with minimal information loss. This bio-inspired design is memory efficient, computationally fast, avoids time-blocking (i.e. fixed and predefined frame rates), and contains "local memory" from past data. The design is evaluated on a wide range of challenging tasks (e.g. event denoising, image reconstruction, classification, and human pose estimation) and is shown to dramatically improve state-of-the-art performance. TORE volumes are an easy-to-implement replacement for any algorithm currently utilizing event representations.

Ruixu Liu, Ju Shen, He Wang et al.

The attention mechanism provides a sequential prediction framework for learning spatial models with enhanced implicit temporal consistency. In this work, we show a systematic design (from 2D to 3D) for how conventional networks and other forms of constraints can be incorporated into the attention framework for learning long-range dependencies for the task of pose estimation. The contribution of this paper is to provide a systematic approach for designing and training of attention-based models for the end-to-end pose estimation, with the flexibility and scalability of arbitrary video sequences as input. We achieve this by adapting temporal receptive field via a multi-scale structure of dilated convolutions. Besides, the proposed architecture can be easily adapted to a causal model enabling real-time performance. Any off-the-shelf 2D pose estimation systems, e.g. Mocap libraries, can be easily integrated in an ad-hoc fashion. Our method achieves the state-of-the-art performance and outperforms existing methods by reducing the mean per joint position error to 33.4 mm on Human3.6M dataset.

Almabrok Essa, Vijayan Asari

Derived from a general definition of texture in a local neighborhood, local directional pattern (LDP) encodes the directional information in the small local 3x3 neighborhood of a pixel, which may fail to extract detailed information especially during changes in the input image due to illumination variations. Therefore, in this paper we introduce a novel feature extraction technique that calculates the nth order direction variation patterns, named high order local directional pattern (HOLDP). The proposed HOLDP can capture more detailed discriminative information than the conventional LDP. Unlike the LDP operator, our proposed technique extracts nth order local information by encoding various distinctive spatial relationships from each neighborhood layer of a pixel in the pyramidal multi-structure way. Then we concatenate the feature vector of each neighborhood layer to form the final HOLDP feature vector. The performance evaluation of the proposed HOLDP algorithm is conducted on several publicly available face databases and observed the superiority of HOLDP under extreme illumination conditions.

Nina Varney, Vijayan K. Asari, Quinn Graehling

We present a method to learn a diverse group of object categories from an unordered point set. We propose our Pyramid Point network, which uses a dense pyramid structure instead of the traditional 'U' shape, typically seen in semantic segmentation networks. This pyramid structure gives a second look, allowing the network to revisit different layers simultaneously, increasing the contextual information by creating additional layers with less noise. We introduce a Focused Kernel Point convolution (FKP Conv), which expands on the traditional point convolutions by adding an attention mechanism to the kernel outputs. This FKP Conv increases our feature quality and allows us to weigh the kernel outputs dynamically. These FKP Convs are the central part of our Recurrent FKP Bottleneck block, which makes up the backbone of our encoder. With this distinct network, we demonstrate competitive performance on three benchmark data sets. We also perform an ablation study to show the positive effects of each element in our FKP Conv.

Md Zahangir Alom, Raj P. Kapur, TJ Browen et al.

Hirschsprungs disease (HD) is a birth defect which is diagnosed and managed by multiple medical specialties such as pediatric gastroenterology, surgery, radiology, and pathology. HD is characterized by absence of ganglion cells in the distal intestinal tract with a gradual normalization of ganglion cell numbers in adjacent upstream bowel, termed as the transition zone (TZ). Definitive surgical management to remove the abnormal bowel requires accurate assessment of ganglion cell density in histological sections from the TZ, which is difficult, time-consuming and prone to operator error. We present an automated method to detect and count immunostained ganglion cells using a new NABLA_N network based deep learning (DL) approach, called GanglionNet. The morphological image analysis methods are applied for refinement of the regions for counting of the cells and define ganglia regions (a set of ganglion cells) from the predicted masks. The proposed model is trained with single point annotated samples by the expert pathologist. The GanglionNet is tested on ten completely new High Power Field (HPF) images with dimension of 2560x1920 pixels and the outputs are compared against the manual counting results by the expert pathologist. The proposed method shows a robust 97.49% detection accuracy for ganglion cells, when compared to counts by the expert pathologist, which demonstrates the robustness of GanglionNet. The proposed DL based ganglion cell detection and counting method will simplify and standardize TZ diagnosis for HD patients.

Nina Varney, Vijayan K. Asari, Quinn Graehling

We present the Dayton Annotated LiDAR Earth Scan (DALES) data set, a new large-scale aerial LiDAR data set with over a half-billion hand-labeled points spanning 10 square kilometers of area and eight object categories. Large annotated point cloud data sets have become the standard for evaluating deep learning methods. However, most of the existing data sets focus on data collected from a mobile or terrestrial scanner with few focusing on aerial data. Point cloud data collected from an Aerial Laser Scanner (ALS) presents a new set of challenges and applications in areas such as 3D urban modeling and large-scale surveillance. DALES is the most extensive publicly available ALS data set with over 400 times the number of points and six times the resolution of other currently available annotated aerial point cloud data sets. This data set gives a critical number of expert verified hand-labeled points for the evaluation of new 3D deep learning algorithms, helping to expand the focus of current algorithms to aerial data. We describe the nature of our data, annotation workflow, and provide a benchmark of current state-of-the-art algorithm performance on the DALES data set.

Md Zahangir Alom, M M Shaifur Rahman, Mst Shamima Nasrin et al.

COVID-19 is currently one the most life-threatening problems around the world. The fast and accurate detection of the COVID-19 infection is essential to identify, take better decisions and ensure treatment for the patients which will help save their lives. In this paper, we propose a fast and efficient way to identify COVID-19 patients with multi-task deep learning (DL) methods. Both X-ray and CT scan images are considered to evaluate the proposed technique. We employ our Inception Residual Recurrent Convolutional Neural Network with Transfer Learning (TL) approach for COVID-19 detection and our NABLA-N network model for segmenting the regions infected by COVID-19. The detection model shows around 84.67% testing accuracy from X-ray images and 98.78% accuracy in CT-images. A novel quantitative analysis strategy is also proposed in this paper to determine the percentage of infected regions in X-ray and CT images. The qualitative and quantitative results demonstrate promising results for COVID-19 detection and infected region localization.

R. Wes Baldwin, Mohammed Almatrafi, Vijayan Asari et al.

This paper presents a novel method for labeling real-world neuromorphic camera sensor data by calculating the likelihood of generating an event at each pixel within a short time window, which we refer to as "event probability mask" or EPM. Its applications include (i) objective benchmarking of event denoising performance, (ii) training convolutional neural networks for noise removal called "event denoising convolutional neural network" (EDnCNN), and (iii) estimating internal neuromorphic camera parameters. We provide the first dataset (DVSNOISE20) of real-world labeled neuromorphic camera events for noise removal.

Ming Gong, Liping Yang, Catherine Potts et al.

Line segment detection is an essential task in computer vision and image analysis, as it is the critical foundation for advanced tasks such as shape modeling and road lane line detection for autonomous driving. We present a robust topological graph guided approach for line segment detection in low quality binary images (hence, we call it TGGLines). Due to the graph-guided approach, TGGLines not only detects line segments, but also organizes the segments with a line segment connectivity graph, which means the topological relationships (e.g., intersection, an isolated line segment) of the detected line segments are captured and stored; whereas other line detectors only retain a collection of loose line segments. Our empirical results show that the TGGLines detector visually and quantitatively outperforms state-of-the-art line segment detection methods. In addition, our TGGLines approach has the following two competitive advantages: (1) our method only requires one parameter and it is adaptive, whereas almost all other line segment detection methods require multiple (non-adaptive) parameters, and (2) the line segments detected by TGGLines are organized by a line segment connectivity graph.

R Wes Baldwin, Mohammed Almatrafi, Jason R Kaufman et al.

This paper presents a novel fusion of low-level approaches for dimensionality reduction into an effective approach for high-level objects in neuromorphic camera data called Inceptive Event Time-Surfaces (IETS). IETSs overcome several limitations of conventional time-surfaces by increasing robustness to noise, promoting spatial consistency, and improving the temporal localization of (moving) edges. Combining IETS with transfer learning improves state-of-the-art performance on the challenging problem of object classification utilizing event camera data.

Md Zahangir Alom, Theus Aspiras, Tarek M. Taha et al.

In the last few years, Deep Learning (DL) has been showing superior performance in different modalities of biomedical image analysis. Several DL architectures have been proposed for classification, segmentation, and detection tasks in medical imaging and computational pathology. In this paper, we propose a new DL architecture, the NABLA-N network, with better feature fusion techniques in decoding units for dermoscopic image segmentation tasks. The NABLA-N network has several advances for segmentation tasks. First, this model ensures better feature representation for semantic segmentation with a combination of low to high-level feature maps. Second, this network shows better quantitative and qualitative results with the same or fewer network parameters compared to other methods. In addition, the Inception Recurrent Residual Convolutional Neural Network (IRRCNN) model is used for skin cancer classification. The proposed NABLA-N network and IRRCNN models are evaluated for skin cancer segmentation and classification on the benchmark datasets from the International Skin Imaging Collaboration 2018 (ISIC-2018). The experimental results show superior performance on segmentation tasks compared to the Recurrent Residual U-Net (R2U-Net). The classification model shows around 87% testing accuracy for dermoscopic skin cancer classification on ISIC2018 dataset.

Md Zahangir Alom, Theus Aspiras, Tarek M. Taha et al.

Deep Learning (DL) approaches have been providing state-of-the-art performance in different modalities in the field of medical imagining including Digital Pathology Image Analysis (DPIA). Out of many different DL approaches, Deep Convolutional Neural Network (DCNN) technique provides superior performance for classification, segmentation, and detection tasks. Most of the task in DPIA problems are somehow possible to solve with classification, segmentation, and detection approaches. In addition, sometimes pre and post-processing methods are applied for solving some specific type of problems. Recently, different DCNN models including Inception residual recurrent CNN (IRRCNN), Densely Connected Recurrent Convolution Network (DCRCN), Recurrent Residual U-Net (R2U-Net), and R2U-Net based regression model (UD-Net) have proposed and provide state-of-the-art performance for different computer vision and medical image analysis tasks. However, these advanced DCNN models have not been explored for solving different problems related to DPIA. In this study, we have applied these DCNN techniques for solving different DPIA problems and evaluated on different publicly available benchmark datasets for seven different tasks in digital pathology including lymphoma classification, Invasive Ductal Carcinoma (IDC) detection, nuclei segmentation, epithelium segmentation, tubule segmentation, lymphocyte detection, and mitosis detection. The experimental results are evaluated with different performance metrics such as sensitivity, specificity, accuracy, F1-score, Receiver Operating Characteristics (ROC) curve, dice coefficient (DC), and Means Squired Errors (MSE). The results demonstrate superior performance for classification, segmentation, and detection tasks compared to existing machine learning and DCNN based approaches.

Md Zahangir Alom, Chris Yakopcic, Tarek M. Taha et al.

The Deep Convolutional Neural Network (DCNN) is one of the most powerful and successful deep learning approaches. DCNNs have already provided superior performance in different modalities of medical imaging including breast cancer classification, segmentation, and detection. Breast cancer is one of the most common and dangerous cancers impacting women worldwide. In this paper, we have proposed a method for breast cancer classification with the Inception Recurrent Residual Convolutional Neural Network (IRRCNN) model. The IRRCNN is a powerful DCNN model that combines the strength of the Inception Network (Inception-v4), the Residual Network (ResNet), and the Recurrent Convolutional Neural Network (RCNN). The IRRCNN shows superior performance against equivalent Inception Networks, Residual Networks, and RCNNs for object recognition tasks. In this paper, the IRRCNN approach is applied for breast cancer classification on two publicly available datasets including BreakHis and Breast Cancer Classification Challenge 2015. The experimental results are compared against the existing machine learning and deep learning-based approaches with respect to image-based, patch-based, image-level, and patient-level classification. The IRRCNN model provides superior classification performance in terms of sensitivity, Area Under the Curve (AUC), the ROC curve, and global accuracy compared to existing approaches for both datasets.

Md Zahangir Alom, Chris Yakopcic, Tarek M. Taha et al.

Due to cellular heterogeneity, cell nuclei classification, segmentation, and detection from pathological images are challenging tasks. In the last few years, Deep Convolutional Neural Networks (DCNN) approaches have been shown state-of-the-art (SOTA) performance on histopathological imaging in different studies. In this work, we have proposed different advanced DCNN models and evaluated for nuclei classification, segmentation, and detection. First, the Densely Connected Recurrent Convolutional Network (DCRN) model is used for nuclei classification. Second, Recurrent Residual U-Net (R2U-Net) is applied for nuclei segmentation. Third, the R2U-Net regression model which is named UD-Net is used for nuclei detection from pathological images. The experiments are conducted with different datasets including Routine Colon Cancer(RCC) classification and detection dataset, and Nuclei Segmentation Challenge 2018 dataset. The experimental results show that the proposed DCNN models provide superior performance compared to the existing approaches for nuclei classification, segmentation, and detection tasks. The results are evaluated with different performance metrics including precision, recall, Dice Coefficient (DC), Means Squared Errors (MSE), F1-score, and overall accuracy. We have achieved around 3.4% and 4.5% better F-1 score for nuclei classification and detection tasks compared to recently published DCNN based method. In addition, R2U-Net shows around 92.15% testing accuracy in term of DC. These improved methods will help for pathological practices for better quantitative analysis of nuclei in Whole Slide Images(WSI) which ultimately will help for better understanding of different types of cancer in clinical workflow.

Md Zahangir Alom, Tarek M. Taha, Christopher Yakopcic et al.

Deep learning has demonstrated tremendous success in variety of application domains in the past few years. This new field of machine learning has been growing rapidly and applied in most of the application domains with some new modalities of applications, which helps to open new opportunity. There are different methods have been proposed on different category of learning approaches, which includes supervised, semi-supervised and un-supervised learning. The experimental results show state-of-the-art performance of deep learning over traditional machine learning approaches in the field of Image Processing, Computer Vision, Speech Recognition, Machine Translation, Art, Medical imaging, Medical information processing, Robotics and control, Bio-informatics, Natural Language Processing (NLP), Cyber security, and many more. This report presents a brief survey on development of DL approaches, including Deep Neural Network (DNN), Convolutional Neural Network (CNN), Recurrent Neural Network (RNN) including Long Short Term Memory (LSTM) and Gated Recurrent Units (GRU), Auto-Encoder (AE), Deep Belief Network (DBN), Generative Adversarial Network (GAN), and Deep Reinforcement Learning (DRL). In addition, we have included recent development of proposed advanced variant DL techniques based on the mentioned DL approaches. Furthermore, DL approaches have explored and evaluated in different application domains are also included in this survey. We have also comprised recently developed frameworks, SDKs, and benchmark datasets that are used for implementing and evaluating deep learning approaches. There are some surveys have published on Deep Learning in Neural Networks [1, 38] and a survey on RL [234]. However, those papers have not discussed the individual advanced techniques for training large scale deep learning models and the recently developed method of generative models [1].

Md Zahangir Alom, Mahmudul Hasan, Chris Yakopcic et al.

Deep learning (DL) based semantic segmentation methods have been providing state-of-the-art performance in the last few years. More specifically, these techniques have been successfully applied to medical image classification, segmentation, and detection tasks. One deep learning technique, U-Net, has become one of the most popular for these applications. In this paper, we propose a Recurrent Convolutional Neural Network (RCNN) based on U-Net as well as a Recurrent Residual Convolutional Neural Network (RRCNN) based on U-Net models, which are named RU-Net and R2U-Net respectively. The proposed models utilize the power of U-Net, Residual Network, as well as RCNN. There are several advantages of these proposed architectures for segmentation tasks. First, a residual unit helps when training deep architecture. Second, feature accumulation with recurrent residual convolutional layers ensures better feature representation for segmentation tasks. Third, it allows us to design better U-Net architecture with same number of network parameters with better performance for medical image segmentation. The proposed models are tested on three benchmark datasets such as blood vessel segmentation in retina images, skin cancer segmentation, and lung lesion segmentation. The experimental results show superior performance on segmentation tasks compared to equivalent models including U-Net and residual U-Net (ResU-Net).

Md Zahangir Alom, Mahmudul Hasan, Chris Yakopcic et al.

Machine learning and computer vision have driven many of the greatest advances in the modeling of Deep Convolutional Neural Networks (DCNNs). Nowadays, most of the research has been focused on improving recognition accuracy with better DCNN models and learning approaches. The recurrent convolutional approach is not applied very much, other than in a few DCNN architectures. On the other hand, Inception-v4 and Residual networks have promptly become popular among computer the vision community. In this paper, we introduce a new DCNN model called the Inception Recurrent Residual Convolutional Neural Network (IRRCNN), which utilizes the power of the Recurrent Convolutional Neural Network (RCNN), the Inception network, and the Residual network. This approach improves the recognition accuracy of the Inception-residual network with same number of network parameters. In addition, this proposed architecture generalizes the Inception network, the RCNN, and the Residual network with significantly improved training accuracy. We have empirically evaluated the performance of the IRRCNN model on different benchmarks including CIFAR-10, CIFAR-100, TinyImageNet-200, and CU3D-100. The experimental results show higher recognition accuracy against most of the popular DCNN models including the RCNN. We have also investigated the performance of the IRRCNN approach against the Equivalent Inception Network (EIN) and the Equivalent Inception Residual Network (EIRN) counterpart on the CIFAR-100 dataset. We report around 4.53%, 4.49% and 3.56% improvement in classification accuracy compared with the RCNN, EIN, and EIRN on the CIFAR-100 dataset respectively. Furthermore, the experiment has been conducted on the TinyImageNet-200 and CU3D-100 datasets where the IRRCNN provides better testing accuracy compared to the Inception Recurrent CNN (IRCNN), the EIN, and the EIRN.

Md Zahangir Alom, Peheding Sidike, Mahmudul Hasan et al.

In spite of advances in object recognition technology, Handwritten Bangla Character Recognition (HBCR) remains largely unsolved due to the presence of many ambiguous handwritten characters and excessively cursive Bangla handwritings. Even the best existing recognizers do not lead to satisfactory performance for practical applications related to Bangla character recognition and have much lower performance than those developed for English alpha-numeric characters. To improve the performance of HBCR, we herein present the application of the state-of-the-art Deep Convolutional Neural Networks (DCNN) including VGG Network, All Convolution Network (All-Conv Net), Network in Network (NiN), Residual Network, FractalNet, and DenseNet for HBCR. The deep learning approaches have the advantage of extracting and using feature information, improving the recognition of 2D shapes with a high degree of invariance to translation, scaling and other distortions. We systematically evaluated the performance of DCNN models on publicly available Bangla handwritten character dataset called CMATERdb and achieved the superior recognition accuracy when using DCNN models. This improvement would help in building an automatic HBCR system for practical applications.

Md Zahangir Alom, Paheding Sidike, Tarek M. Taha et al.

In spite of the advances in pattern recognition technology, Handwritten Bangla Character Recognition (HBCR) (such as alpha-numeric and special characters) remains largely unsolved due to the presence of many perplexing characters and excessive cursive in Bangla handwriting. Even the best existing recognizers do not lead to satisfactory performance for practical applications. To improve the performance of Handwritten Bangla Digit Recognition (HBDR), we herein present a new approach based on deep neural networks which have recently shown excellent performance in many pattern recognition and machine learning applications, but has not been throughly attempted for HBDR. We introduce Bangla digit recognition techniques based on Deep Belief Network (DBN), Convolutional Neural Networks (CNN), CNN with dropout, CNN with dropout and Gaussian filters, and CNN with dropout and Gabor filters. These networks have the advantage of extracting and using feature information, improving the recognition of two dimensional shapes with a high degree of invariance to translation, scaling and other pattern distortions. We systematically evaluated the performance of our method on publicly available Bangla numeral image database named CMATERdb 3.1.1. From experiments, we achieved 98.78% recognition rate using the proposed method: CNN with Gabor features and dropout, which outperforms the state-of-the-art algorithms for HDBR.

Kumar S. Ray, Vijayan K. Asari, Soma Chakraborty

In this paper we propose a novel approach for detecting and tracking objects in videos with variable background i.e. videos captured by moving cameras without any additional sensor. In a video captured by a moving camera, both the background and foreground are changing in each frame of the image sequence. So for these videos, modeling a single background with traditional background modeling methods is infeasible and thus the detection of actual moving object in a variable background is a challenging task. To detect actual moving object in this work, spatio-temporal blobs have been generated in each frame by spatio-temporal analysis of the image sequence using a three-dimensional Gabor filter. Then individual blobs, which are parts of one object are merged using Minimum Spanning Tree to form the moving object in the variable background. The height, width and four-bin gray-value histogram of the object are calculated as its features and an object is tracked in each frame using these features to generate the trajectories of the object through the video sequence. In this work, problem of data association during tracking is solved by Linear Assignment Problem and occlusion is handled by the application of kalman filter. The major advantage of our method over most of the existing tracking algorithms is that, the proposed method does not require initialization in the first frame or training on sample data to perform. Performance of the algorithm has been tested on benchmark videos and very satisfactory result has been achieved. The performance of the algorithm is also comparable and superior with respect to some benchmark algorithms.