Keypoints Detection and Feature Extraction: A Dynamic Genetic Programming Approach for Evolving Rotation-invariant Texture Image Descriptors

The goodness of the features extracted from the instances and the number of training instances are two key components in machine learning, and building an effective model is largely affected by these two factors. Acquiring a large number of training instances is very expensive in some situations such as in the medical domain. Designing a good feature set, on the other hand, is very hard and often requires domain expertise. In computer vision, image descriptors have emerged to automate feature detection and extraction; however, domain-expert intervention is typically needed to develop these descriptors. The aim of this paper is to utilize genetic programming to automatically construct a rotation-invariant image descriptor by synthesizing a set of formulas using simple arithmetic operators and first-order statistics, and determining the length of the feature vector simultaneously using only two instances per class. Using seven texture classification image datasets, the performance of the proposed method is evaluated and compared against eight domain-expert hand-crafted image descriptors. Quantitatively, the proposed method has significantly outperformed, or achieved comparable performance to, the competitor methods. Qualitatively, the analysis shows that the descriptors evolved by the proposed method can be interpreted.

Feature extraction, as one essential step of image classification, can potentially reduce image data dimensionality and capture effective information for improving performance. However, most existing image descriptors are designed to conduct specific tasks and might not be sufficient for different types of images. Genetic programming (GP) can automatically extract multiple important and discriminative features by incorporating diverse image descriptors into a GP program. Furthermore, different regions in an image have different structural characteristics. In this paper, we propose a region adaptive image classification approach based on GP, which can automatically extract informative image features by automatically applying different image descriptors in different regions of an image. A new flexible GP program structure with a new function set and a new terminal set is developed in this approach. The performance of the proposed method is evaluated on four various data sets and compared with other state-of-the-art classification methods. Experimental results illustrate that the proposed approach is capable of achieving better or competitive performance than these baseline methods. Further analysis of some good programs shows the high interpretability of the proposed method.

COVID-19 is highly infectious and causes acute respiratory disease. Machine learning (ML) and deep learning (DL) models are vital in detecting disease from computerized chest tomography (CT) scans. The DL models outperformed the ML models. For COVID-19 detection from CT scan images, DL models are used as end-to-end models. Thus, the performance of the model is evaluated for the quality of the extracted feature and classification accuracy. There are four contributions included in this work. First, this research is motivated by studying the quality of the extracted feature from the DL by feeding these extracted to an ML model. In other words, we proposed comparing the end-to-end DL model performance against the approach of using DL for feature extraction and ML for the classification of COVID-19 CT scan images. Second, we proposed studying the effect of fusing extracted features from image descriptors, e.g., Scale-Invariant Feature Transform (SIFT), with extracted features from DL models. Third, we proposed a new Convolutional Neural Network (CNN) to be trained from scratch and then compared to the deep transfer learning on the same classification problem. Finally, we studied the performance gap between classic ML models against ensemble learning models. The proposed framework is evaluated using a CT dataset, where the obtained results are evaluated using five different metrics The obtained results revealed that using the proposed CNN model is better than using the well-known DL model for the purpose of feature extraction. Moreover, using a DL model for feature extraction and an ML model for the classification task achieved better results in comparison to using an end-to-end DL model for detecting COVID-19 CT scan images. Of note, the accuracy rate of the former method improved by using ensemble learning models instead of the classic ML models. The proposed method achieved the best accuracy rate of 99.39%.

This paper reviews objective methods for prognostic modelling of cancer tumours located within radiology images, a process known as radiomics. Radiomics is a novel feature transformation method for detecting clinically relevant features from radiological imaging data that are difficult for the human eye to perceive. To facilitate the detection machine learning and deep learning methods are increasingly investigated with the aim of improving patient diagnosis, treatment options and outcomes. A review of the relevant works in the expanding field of radiomics for survival prediction from cancer is provided. Research works outside the field of radiomics which define techniques that may be of future use to improve feature extraction and analysis are also reviewed. Radiomics is a rapidly advancing field of clinical image analysis with a vast potential for supporting decision making involved in the diagnosis and treatment of cancer. The realisation of this goal of more effective decision making requires significant individual and integrated expertise from domain experts in medicine, biology and computer science to allow advances in computer vision and machine learning techniques to be applied effectively. Deep learning combined with machine learning has the potential to advance the field of radiomics significantly in the years to come, provided that mechanisms for data sharing or distributed learning are established to increase the availability of data across all patient and tumour types.

Composite fault diagnosis of analog circuit system using chaotic game optimization-assisted deep ELM-AE

Bag of Features (BoF) has been successfully exploited in content-based image retrieval for several years. Due to its performance and popularity, several extensions have been proposed that involve feature description, dictionary building, feature encoding and post-query process, etc. This paper proposes a dynamically weighting scheme for BoF-based image retrieval based on feedback. It involves two contributions: (i) analyzing the statistical distribution characteristic of similar BoF representations and (ii) computing weights dynamically based on the feedback obtained from different initial query results. We quantitatively evaluate the proposed method on two different databases. Experiments confirm that the proposed weighting scheme has better performance than the baseline of BoF-based image retrieval systems. Meanwhile, the results demonstrate the effectiveness of the weighting scheme in terms of the precision of top-N returned images.

A modified genetic algorithm and weighted principal component analysis based feature selection and extraction strategy in agriculture

Machine Learning (ML) applications in complex domains, such as the medical domain, can be highly beneficial, but also hazardous if some concepts are overlooked. In this context, however, health professionals denote expertise in their domain, but they often lack skills in terms of ML. In this sense, to leverage ML applications in the medical domain, it is important to combine both domain expertise and ML-related skills. In previous works, we tackled this challenge in the health context through a visual platform (KoopaML) that enables lay users to build ML pipelines. The present work describes the challenges derived from the first version of the platform and the prototypes for the new features designed to address them. The prototypes have been validated by two experts, obtaining highly valuable feedback.

BackgroundMachine learning, especially deep learning, is becoming more and more relevant in research and development in the medical domain. For all the supervised deep learning applications, data is the most critical factor in securing successful implementation and sustaining the progress of the machine learning model. Especially gastroenterological data, which often involves endoscopic videos, are cumbersome to annotate. Domain experts are needed to interpret and annotate the videos. To support those domain experts, we generated a framework. With this framework, instead of annotating every frame in the video sequence, experts are just performing key annotations at the beginning and the end of sequences with pathologies, e.g., visible polyps. Subsequently, non-expert annotators supported by machine learning add the missing annotations for the frames in-between.MethodsIn our framework, an expert reviews the video and annotates a few video frames to verify the object’s annotations for the non-expert. In a second step, a non-expert has visual confirmation of the given object and can annotate all following and preceding frames with AI assistance. After the expert has finished, relevant frames will be selected and passed on to an AI model. This information allows the AI model to detect and mark the desired object on all following and preceding frames with an annotation. Therefore, the non-expert can adjust and modify the AI predictions and export the results, which can then be used to train the AI model.ResultsUsing this framework, we were able to reduce workload of domain experts on average by a factor of 20 on our data. This is primarily due to the structure of the framework, which is designed to minimize the workload of the domain expert. Pairing this framework with a state-of-the-art semi-automated AI model enhances the annotation speed further. Through a prospective study with 10 participants, we show that semi-automated annotation using our tool doubles the annotation speed of non-expert annotators compared to a well-known state-of-the-art annotation tool.ConclusionIn summary, we introduce a framework for fast expert annotation for gastroenterologists, which reduces the workload of the domain expert considerably while maintaining a very high annotation quality. The framework incorporates a semi-automated annotation system utilizing trained object detection models. The software and framework are open-source.

ISAR imagery is quite different from optical images. Thus, feature extraction from 2-D radar imagery is more complicated. In this chapter, we introduced two distinct instances of feature extraction in ISAR: 2-D feature extraction from a normal ISAR range-Doppler (or cross-range) image; and target motion feature extraction from micro-Doppler signatures. For 2-D feature extraction, image feature extraction in computer vision (e.g., the model-based method, morphology method, template matching, and machine learning) may be used. In Section 7.1, we introduced how to estimate the size and orientation of a target from its range and cross-range extents and how to estimate the target rotation parameters based on the estimated target dimension. In Section 7.2, we introduced how to extract target motion features showing in ISAR images, which are extracted from ISAR range profiles instead of 2-D ISAR imagery. Time-varying micro-Doppler signatures provide information that allows the estimation of various kinematic features. We demonstrated these with helicopter rotor blades and simulated walking human data.

There is a rich literature of approaches to image feature extraction in computer vision. Many sophisticated approaches exist for low- and high-level feature extraction but can be complex to implement with parameter choice guided by experimentation, but impeded by speed of computation. We have developed new ways to extract features based on notional use of physical paradigms, with parameterisation that is more familiar to a scientifically-trained user, aiming to make best use of computational resource. We describe how analogies based on gravitational force can be used for low-level analysis, whilst analogies of water flow and heat can be deployed to achieve high-level smooth shape detection. These new approaches to arbitrary shape extraction are compared with standard state-of-art approaches by curve evolution. There is no comparator operator to our use of gravitational force. We also aim to show that the implementation is consistent with the original motivations for these techniques and so contend that the exploration of physical paradigms offers a promising new avenue for new approaches to feature extraction in computer vision.

Distinguishing malicious programs based on visualization and hybrid learning algorithms

Principal curves are nonlinear generalizations of principal components analysis. They are smooth self-consistent curves that pass through the middle of the distribution. By analysis of existed principal curves, we learn that a soft k-segments algorithm for principal curves exhibits good performance in such situations in which the data sets are concentrated around a highly curved or self-intersecting curves. Extraction of features are critical to improve the recognition rate of off-line handwritten characters. Therefore, we attempt to use the algorithm to extract structural features of off-line handwritten characters. Experiment results show that the algorithm is not only feasible for extraction of structural features of characters, but also exhibits good performance. The proposed method can provide a new approach to the research for extraction of structural features of characters.

Characterization of coronary plaque regions in intravascular ultrasound images using a hybrid ensemble classifier

Visual question answering (VQA) is receiving increasing attention from researchers in both the computer vision and natural language processing fields. There are two key components in the VQA task: feature extraction and multi-modal fusion. For feature extraction, we introduce a novel co-attention scheme by combining Sentence-guide Word Attention (SWA) and Question-guide Image Attention in a unified framework. To be specific, the textual attention SWA relies on the semantics of the whole question sentence to calculate contributions of different question words for text representation. For the multi-modal fusion, we propose a “Cross-modal Multistep Fusion (CMF)” network to generate multistep features and achieve multiple interactions for two modalities, rather than focusing on modeling complex interactions between two modals like most current feature fusion methods. To avoid the linear increase of the computational cost, we share the parameters for each step in the CMF. Extensive experiments demonstrate that the proposed method can achieve competitive or better performance than the state of the art.