Abstract: The evolution and integration of automated vehicles have become more and more tangible in recent years. State-of-the-art technological advances in the field of camera-based Artificial Intelligence (AI) and computer vision greatly favor the performance and reliability of Advanced Driver Assistance System (ADAS), leading to a greater knowledge of vehicular operation and resembling the human behaviour. However, the exclusive use of this technology still seems insufficient to control the vehicular operation at 100%. To reveal the degree of accuracy of the current camera-based automated driving AI modules, this paper studies the structure and behavior of one of the main solutions in a controlled testing environment. The results obtained clearly outline the lack of reliability when using exclusively the AI model in the perception stage, thereby entailing using additional complementary sensors to improve its safety and performance.
Abstract: Medical methods to completely treat glaucoma are yet to be developed. Therefore, ophthalmologists manage patients mainly to delay disease progression. Patients with glaucoma are mainly elderly individuals. In elderly people's houses, having an equipment that can provide medical treatment and care can release their family from their care. For elderly people with the glaucoma to live by themselves as much as possible, we developed a support robot having five functions: elderly people care, ophthalmological examination, trip assistance to the neighborhood, medical treatment, and data referral to a hospital. The medical and nursing care robot should approach the visual field that the patients can see at a speed suitable for their eyesight. This is because the robot will be dangerous if it approaches the patients from the visual field that they cannot see. We experimentally developed a robot that brings a white cane to elderly people with glaucoma. The base part of the robot is a carriage, which is a Megarover 1.1, and it has two infrared sensors. The robot moves along a white line on the floor using the infrared sensors and has a special arm, which does not use electricity. The arm can scoop the block attached to the white cane. Next, we also developed a direction detector comprised of a charge-coupled device camera (SVR41ResucueHD; Sun Mechatronics), goggles (MG-277MLF; Midori Anzen Co. Ltd.), and biconvex lenses with a focal length of 25 mm (Edmund Co.). Some young people were photographed using the direction detector, which was put on their faces. Image processing was performed using Scilab 6.1.0 and Image Processing and Computer Vision Toolbox 4.1.2. To measure the people's line of vision, we calculated the iris's center of gravity using five processes: reduction, trimming, binarization or gray scale, edge extraction, and Hough transform. We compared the binarization and gray scale processes in image processing. The binarization process was better than the gray scale process. For edge extraction, we compared five methods: Sobel, Prewitt, Laplacian of Gaussian, fast Fourier transform, and Canny. The Canny method was the optimal extraction method. We performed the Hough transform to search for the main coordinates from the iris's edge, and we found that the Hough transform could calculate the center point of the iris.
Abstract: Depth estimation using stereo images is a challenging problem in computer vision. Many different studies have been carried out to solve this problem. With advancing machine learning, tackling this problem is often done with neural network-based solutions. The images used in these studies are mostly in the visible spectrum. However, the need to use the Infrared (IR) spectrum for depth estimation has emerged because it gives better results than visible spectra in some conditions. At this point, we recommend using thermal-thermal (IR) image pairs for depth estimation. In this study, we used two well-known networks (PSMNet, FADNet) with minor modifications to demonstrate the viability of this idea.
Abstract: In this paper, we predict the likelihood of a player making a shot in basketball from multiagent trajectories. To approach this problem, we present a convolutional neural network (CNN) approach where we initially represent the multiagent behavior as an image. To encode the adversarial nature of basketball, we use a multichannel image which we then feed into a CNN. Additionally, to capture the temporal aspect of the trajectories we use “fading.” We find that this approach is superior to a traditional FFN model. By using gradient ascent, we were able to discover what the CNN filters look for during training. Last, we find that a combined FFN+CNN is the best performing network with an error rate of 39%.
Abstract: The purpose of this study is to explore the characteristics of developing a machine learning application using synthetic data. The study is structured to develop the application for the purpose of deploying the computer vision model. The findings discuss the realities of attempting to develop a computer vision model for practical purpose, and detail the processes, tools and techniques that were used to meet accuracy requirements. The research reveals that synthetic data represent another variable that can be adjusted to improve the performance of a computer vision model. Further, a suite of tools and tuning recommendations are provided.
Abstract: Tree-ring analysis is an important part of the quality assessment and the dating of (archaeological) wood samples. It provides quantitative data about the whole anatomical ring structure, which can be used, for example, to measure the impact of the fluctuating environment on the tree growth, for the dendrochronological analysis of archaeological wooden artefacts and to estimate the wood mechanical properties. Despite advances in computer vision and edge recognition algorithms, detection and counting of annual rings are still limited to 2D datasets and performed in most cases manually, which is a time consuming, tedious task and depends strongly on the operator’s experience. This work presents an image processing approach to detect the whole 3D tree-ring structure directly from X-ray computed tomography imaging data. The approach relies on a modified Canny edge detection algorithm, which captures fully connected tree-ring edges throughout the measured image stack and is validated on X-ray computed tomography data taken from six wood species.
Abstract: Machine learning is a new and exciting area of
artificial intelligence nowadays. Machine learning is the most
valuable, time, supervised, and cost-effective approach. It is not a
narrow learning approach; it also includes a wide range of methods
and techniques that can be applied to a wide range of complex realworld
problems and time domains. Biological image classification,
adaptive testing, computer vision, natural language processing, object
detection, cancer detection, face recognition, handwriting
recognition, speech recognition, and many other applications of
machine learning are widely used in research, industry, and
government. Every day, more data are generated, and conventional
machine learning techniques are becoming obsolete as users move to
distributed and real-time operations. By providing fundamental
knowledge of machine learning tools and research opportunities in
the field, the aim of this article is to serve as both a comprehensive
overview and a guide. A diverse set of machine learning resources is
demonstrated and contrasted with the key features in this survey.
Abstract: Face is a non-intrusive strong biometrics for
identification of original and dummy facial by different artificial
means. Face recognition is extremely important in the contexts of
computer vision, psychology, surveillance, pattern recognition,
neural network, content based video processing. The availability of a
widespread face database is crucial to test the performance of these
face recognition algorithms. The openly available face databases
include face images with a wide range of poses, illumination, gestures
and face occlusions but there is no dummy face database accessible in
public domain. This paper presents a face detection algorithm based on
the image segmentation in terms of distance from a fixed point and
template matching methods. This proposed work is having the most
appropriate number of nodal points resulting in most appropriate
outcomes in terms of face recognition and detection. The time taken to
identify and extract distinctive facial features is improved in the range
of 90 to 110 sec. with the increment of efficiency by 3%.
Abstract: We as humans use words with accompanying visual and facial cues to communicate effectively. Classifying facial emotion using computer vision methodologies has been an active research area in the computer vision field. In this paper, we propose a simple method for facial expression recognition that enhances accuracy. We tested our method on the FER-2013 dataset that contains static images. Instead of using Histogram equalization to preprocess the dataset, we used Unsharp Mask to emphasize texture and details and sharpened the edges. We also used ImageDataGenerator from Keras library for data augmentation. Then we used Convolutional Neural Networks (CNN) model to classify the images into 7 different facial expressions, yielding an accuracy of 69.46% on the test set. Our results show that using image preprocessing such as the sharpening technique for a CNN model can improve the performance, even when the CNN model is relatively simple.
Abstract: Neural networks are appealing for many applications since they are able to learn complex non-linear relationships between input and output data. As the number of neurons and layers in a neural network increase, it is possible to represent more complex relationships with automatically extracted features. Nowadays Deep Neural Networks (DNNs) are widely used in Computer Vision problems such as; classification, object detection, segmentation image editing etc. In this work, Facial Emotion Recognition task is performed by proposed Convolutional Neural Network (CNN)-based DNN architecture using FER2013 Dataset. Moreover, the effects of different hyperparameters (activation function, kernel size, initializer, batch size and network size) are investigated and ablation study results for Pooling Layer, Dropout and Batch Normalization are presented.
Abstract: Estimating the 6D pose of objects is a core step for robot bin-picking tasks. The problem is that various objects are usually randomly stacked with heavy occlusion in real applications. In this work, we propose a method to regress 6D poses by predicting three points for each object in the 3D point cloud through deep learning. To solve the ambiguity of symmetric pose, we propose a labeling method to help the network converge better. Based on the predicted pose, an iterative method is employed for pose optimization. In real-world experiments, our method outperforms the classical approach in both precision and recall.
Abstract: Pseudomonas aeruginosa is an opportunistic pathogen that forms surface-associated microbial communities (biofilms) on artificial implant devices and on human tissue. Biofilm infections are difficult to treat with antibiotics, in part, because the bacteria in biofilms are physiologically heterogeneous. One measure of biological heterogeneity in a population of cells is to quantify the cellular concentrations of ribosomes, which can be probed with fluorescently labeled nucleic acids. The fluorescent signal intensity following fluorescence in situ hybridization (FISH) analysis correlates to the cellular level of ribosomes. The goals here are to provide computationally and statistically robust approaches to automatically quantify cellular heterogeneity in biofilms from a large library of epifluorescent microscopy FISH images. In this work, the initial steps were developed toward these goals by developing an automated biofilm detection approach for use with FISH images. The approach allows rapid identification of biofilm regions from FISH images that are counterstained with fluorescent dyes. This methodology provides advances over other computational methods, allowing subtraction of spurious signals and non-biological fluorescent substrata. This method will be a robust and user-friendly approach which will enable users to semi-automatically detect biofilm boundaries and extract intensity values from fluorescent images for quantitative analysis of biofilm heterogeneity.
Abstract: With the recent advance of the deep neural network, we observe new applications of NLP (natural language processing) and CV (computer vision) powered by deep neural networks for processing business documents. However, creating a real-world document processing system needs to integrate several NLP and CV tasks, rather than treating them separately. There is a need to have a unified approach for processing documents containing textual and graphical elements with rich formats, diverse layout arrangement, and distinct semantics. In this paper, a framework that fulfills this unified approach is presented. The framework includes a representation model definition for holding the information generated by various tasks and specifications defining the coordination between these tasks. The framework is a blueprint for building a system that can process documents with rich formats, styles, and multiple types of elements. The flexible and lightweight design of the framework can help build a system for diverse business scenarios, such as contract monitoring and reviewing.
Abstract: In the last few years, Automatic Number Plate Recognition (ANPR) systems have become widely used in the safety, the security, and the commercial aspects. Forethought, several methods and techniques are computing to achieve the better levels in terms of accuracy and real time execution. This paper proposed a computer vision algorithm of Number Plate Localization (NPL) and Characters Segmentation (CS). In addition, it proposed an improved method in Optical Character Recognition (OCR) based on Deep Learning (DL) techniques. In order to identify the number of detected plate after NPL and CS steps, the Convolutional Neural Network (CNN) algorithm is proposed. A DL model is developed using four convolution layers, two layers of Maxpooling, and six layers of fully connected. The model was trained by number image database on the Jetson TX2 NVIDIA target. The accuracy result has achieved 95.84%.
Abstract: Ancient books are significant culture inheritors and their background textures convey the potential history information. However, multi-style texture recovery of ancient books has received little attention. Restricted by insufficient ancient textures and complex handling process, the generation of ancient textures confronts with new challenges. For instance, training without sufficient data usually brings about overfitting or mode collapse, so some of the outputs are prone to be fake. Recently, image generation and style transfer based on deep learning are widely applied in computer vision. Breakthroughs within the field make it possible to conduct research upon multi-style texture recovery of ancient books. Under the circumstances, we proposed a network of layout analysis and image fusion system. Firstly, we trained models by using Deep Convolution Generative against Networks (DCGAN) to synthesize multi-style ancient textures; then, we analyzed layouts based on the Position Rearrangement (PR) algorithm that we proposed to adjust the layout structure of foreground content; at last, we realized our goal by fusing rearranged foreground texts and generated background. In experiments, diversified samples such as ancient Yi, Jurchen, Seal were selected as our training sets. Then, the performances of different fine-turning models were gradually improved by adjusting DCGAN model in parameters as well as structures. In order to evaluate the results scientifically, cross entropy loss function and Fréchet Inception Distance (FID) are selected to be our assessment criteria. Eventually, we got model M8 with lowest FID score. Compared with DCGAN model proposed by Radford at el., the FID score of M8 improved by 19.26%, enhancing the quality of the synthetic images profoundly.
Abstract: Human pose estimation and tracking are to accurately identify and locate the positions of human joints in the video. It is a computer vision task which is of great significance for human motion recognition, behavior understanding and scene analysis. There has been remarkable progress on human pose estimation in recent years. However, more researches are needed for human pose tracking especially for online tracking. In this paper, a framework, called PoseSRPN, is proposed for online single-person pose estimation and tracking. We use Siamese network attaching a pose estimation branch to incorporate Single-person Pose Tracking (SPT) and Visual Object Tracking (VOT) into one framework. The pose estimation branch has a simple network structure that replaces the complex upsampling and convolution network structure with deconvolution. By augmenting the loss of fully convolutional Siamese network with the pose estimation task, pose estimation and tracking can be trained in one stage. Once trained, PoseSRPN only relies on a single bounding box initialization and producing human joints location. The experimental results show that while maintaining the good accuracy of pose estimation on COCO and PoseTrack datasets, the proposed method achieves a speed of 59 frame/s, which is superior to other pose tracking frameworks.
Abstract: In this paper we present the design and the implementation of a target tracking system where the target is set to be a moving person in a video sequence. The system can be applied easily as a vision system for mobile robot. The system is composed of two major parts the first is the detection of the person in the video frame using the SVM learning machine based on the “HOG” descriptors. The second part is the tracking of a moving person it’s done by using a combination of the Kalman filter and a modified version of the Camshift tracking algorithm by adding the target motion feature to the color feature, the experimental results had shown that the new algorithm had overcame the traditional Camshift algorithm in robustness and in case of occlusion.
Abstract: Geometric modeling plays an important role in the
constructions and manufacturing of curve, surface and solid
modeling. Their algorithms are critically important not only in
the automobile, ship and aircraft manufacturing business, but are
also absolutely necessary in a wide variety of modern applications,
e.g., robotics, optimization, computer vision, data analytics and
visualization. The calculation and display of geometric objects
can be accomplished by these six techniques: Polynomial basis,
Recursive, Iterative, Coefficient matrix, Polar form approach and
Pyramidal algorithms. In this research, the coefficient matrix (simply
called monomial form approach) will be used to model polynomial
rectangular patches, i.e., Said-Ball, Wang-Ball, DP, Dejdumrong and
NB1 surfaces. Some examples of the monomial forms for these
surface modeling are illustrated in many aspects, e.g., construction,
derivatives, model transformation, degree elevation and degress
reduction.
Abstract: Due to the sensor technology, video surveillance has become the main way for security control in every big city in the world. Surveillance is usually used by governments for intelligence gathering, the prevention of crime, the protection of a process, person, group or object, or the investigation of crime. Many surveillance systems based on computer vision technology have been developed in recent years. Moving target tracking is the most common task for Unmanned Aerial Vehicle (UAV) to find and track objects of interest in mobile aerial surveillance for civilian applications. The paper is focused on vision-based collision avoidance for UAVs by recurrent neural networks. First, images from cameras on UAV were fused based on deep convolutional neural network. Then, a recurrent neural network was constructed to obtain high-level image features for object tracking and extracting low-level image features for noise reducing. The system distributed the calculation of the whole system to local and cloud platform to efficiently perform object detection, tracking and collision avoidance based on multiple UAVs. The experiments on several challenging datasets showed that the proposed algorithm outperforms the state-of-the-art methods.
Abstract: This paper reviews the different concepts that have been investigated concerning the mechanization of fruit thinning as well as multiple working principles and solutions that have been developed for feature extraction of horticultural products, both in the field and industrial environments. The research should be committed towards selective methods, which inevitably need to incorporate some kinds of sensor technology. Computer vision often comes out as an obvious solution for unstructured detection problems, although leaves despite the chosen point of view frequently occlude fruits. Further research on non-traditional sensors that are capable of object differentiation is needed. Ultrasonic and Near Infrared (NIR) technologies have been investigated for applications related to horticultural produce and show a potential to satisfy this need while simultaneously providing spatial information as time of flight sensors. Light Detection and Ranging (LIDAR) technology also shows a huge potential but it implies much greater costs and the related equipment is usually much larger, making it less suitable for portable devices, which may serve a purpose on smaller unstructured orchards. Portable devices may serve a purpose on these types of orchards. In what concerns sensor methods, on-tree fruit detection, major challenge is to overcome the problem of fruits’ occlusion by leaves and branches. Hence, nontraditional sensors capable of providing some type of differentiation should be investigated.