Microswimmers in patterned environments

Text classification (TC) is the task to automatically classify documents based on learned document features. Many popular TC models use simple occurrence of words in a document as features. They also commonly assume word occurrences to be statistically independent in their design. Although it is obvious that such assumption does not hold in general, these TC models have been robust and efficient in their task. Some recent studies have shown context-sensitive TC approaches, which take into consideration contexts in the form of word co-occurrences, have been able to perform better in general. On the other hand, there have been many studies in the use of complex linguistic or semantic features instead of simple word occurrences as features for information retrieval and classification tasks. While these complex features may intuitively have more relevance to the tasks concerned, results of these studies on their effectiveness have been mixed and not been conclusive. In this paper we present our investigation on the use of some complex linguistic features with context-sensitive TC method. Our experiment results show some potential advantages of such approach.

Emergency evacuation with incomplete information in the presence of obstacles

Understanding the tumor microenvironment and detecting rare circulating tumor cells from blood are two major challenges faced by cancer biologists and oncologists. Both require high sensitivity and accuracy in classifying and isolating single cells in a complex and heterogeneous environment. Classical cell classification and sorting techniques are limited by their reliance on pre-selected cell biomarkers or physical characteristics. Recent breakthroughs in machine learning have achieved unprecedented accuracy across a wide range of image classification problems. We have developed a platform that combines high-resolution imaging of unlabeled cells in microfluidic flow with real-time deep neural network (DNN) based classification and sorting. The DNN classifier was trained on more than 25 million high-resolution cell images of multiple types imaged on the platform. Our model was trained to discriminate among multiple cell classes, including immune cell subtypes, non-small-cell lung cancer cells (NSCLC), hepatocellular carcinomas (HCC), and stromal cells (including endothelial, epithelial, fibroblasts, smooth muscle cells). We then assessed model performance on a separate validation set of cell images, including cell lines not used in the training data. Our classifier accurately identifies NSCLC and HCC against a background of blood cells with an area under the ROC curve (AUC) of > 0.999. In addition we demonstrate the enrichment of NSCLC cells from spike-in mixtures with WBCs or whole blood at concentrations as low as 1:100,000, achieving an enrichment of > 25,000x on multiple cell lines. Using dissociated lung cancer tissue, we demonstrate that our label-free classification of tumor cells closely matches results from both standard flow cytometer analysis and single cell RNA sequencing. Additionally we were able to enrich the tumor cell fraction from dissociated tumor tissue thereby improving the sensitivity of mutation detection and enabling refined downstream single-cell genomic analysis. This work demonstrates that deep learning using high-resolution cell images collected at scale can achieve a high classification accuracy and can enable the label-free isolation of rare cells of interest for a wide range of applications. This system can be used to analyze tumor biopsies and liquid biopsies and has the potential to enable the study of tumor cells and tumor microenvironment with novel dimension and insight. Citation Format: Mahyar Salek, Hou-pu Chou, Prashast Khandelwal, Krishna P. Pant, Thomas J. Musci, Nianzhen Li, Christina Chang, Andreja Jovic, Esther Lee, Stephanie Huang, Jeff Walker, Phuc Nguyen, Kiran Saini, Jeanette Mei, Quillan F. Smith, Maddison Masaeli. Deep learning enables label-free profiling of the tumor microenvironment and enrichment of rare cancer cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 188.

The method of modeling and ordering in wavelet domain is very important to design a successful algorithm of embedded image compression. In this paper, the modeling is limited to "pixel classification," the relationship between wavelet pixels in significance coding. Similarly, the ordering is limited to "pixel sorting," the coding order of wavelet pixels. We use pixel classification and sorting to provide a better understanding of previous works. The image pixels in wavelet domain are classified and sorted, either explicitly or implicitly, for embedded image compression. A new embedded image code is proposed based on a novel pixel classification and sorting (PCAS) scheme in wavelet domain. In PCAS, pixels to be coded are classified into several quantized contexts based on a large context template and sorted based on their estimated significance probabilities. The purpose of pixel classification is to exploit the intraband correlation in wavelet domain. Pixel sorting employs several fractional bit-plane coding passes to improve the rate-distortion performance. The proposed pixel classification and sorting technique is simple, yet effective, producing an embedded image code with excellent compression performance. In addition, our algorithm is able to provide either spatial or quality scalability with flexible complexity.

Understanding the often-heterogeneous land cover in urban areas is critical for, among other things, environmental monitoring, spatial planning, and enforcement. Recently, several earth observation satellites were developed with an enhanced spatial resolution that provides for precise and detailed representations of image objects. Morphological image analysis techniques provide useful tools for extracting spatial features from high-resolution, remotely sensed images. This study investigated the efficacy of mathematical morphological (MM) techniques in the land cover classification of a heterogeneous urban landscape using very high-resolution pan-sharpened Pleiades imagery. Specifically, the study evaluated two morphological profiles (MP) techniques (i.e., concatenation of morphological profiles (CMPs) and multi-morphological profiles (MMPs)) in the classification of a heterogeneous urban land cover. The overall accuracies for CMP were 83.14% and 83.19% over the two study areas. Similarly, the MMP overall accuracies were 84.42% and 84.08% for the two study sites. The study concluded that CMP and MMP can greatly improve the classification of heterogeneous landscapes that typify urban areas by effectively representing the structural landscape information necessary for discriminating related land cover classes. In general, similar and visually acceptable results were produced for land cover classification using either CMP or MMP image analysis techniques

Geographic information processing and visualization on new types of platform such as mobile environment or web environment has been regarded as one of industrial and academic issues. For this, the various approaches for mobile GIS and Web GIS are being carried out on the application level. Motivation of this study is on system architecture and a prototype implementation for 3D mobile and Web 3D application. Platforms and devices for mobile 3D GIS technology manipulating geographic data and information provide some limitation to design and implement: limited CPU and memory, absence or limited performance of graphic accelerators, small size of display panel, and so on. Moreover, in the urban application, 3D spatial information is crucial for complex feature modeling. Core functions of mobile 3D GIS by feature-based approach was designed as some modules: data model and design of spatial features, editing and manipulating of 3D landscape objects, generating of geometrically complex type features, supporting of both database and file system, handling of attributes for 3D objects, texture mapping of complex types of 3D objects and digital elevation model with image library containing high resolution satellite imagery. With these functions, an integrated urban 3D modeling and rendering system was implemented using standard mobile 3D graphic API (application programming interface), OPENGL|ES (Open GL for embedded system), with MS EVC 4.0 MFC. As well as this mobile 3D design and implementation, multiple types of geographic features or objects encoded by GML (geography markup language) for the XML-databases building are taken into account of practical applicability for 3D urban approaches. For this process, it was carried out to implement GML editor software with the main functions: supporting and customizing map style sheet, attribute editing, adding/deleting/manipulating geo-base features in fundamental urban data model as the framework model level and user-defined model for database extension, Web publishing through general web browser, and importing/exporting ESRI-shape file format and data transferring to SVG (scalable vector graphics) for 2D transportation features and X3D (extensible Web 3D) for 3D complex features. These two implementations such as mobile 3D and web 3D can be easily linked in data communication and sharing for the 3D urban modeling dealing with complex types of multiple features on integrated architecture.

Application of image processing and radial basis neural network techniques for ore sorting and ore classification

A new embedded image compression algorithm is proposed, based on progressive Pixel Classification And Sorting (PCAS) in wavelet domain. To exploit the intraband and interband correlation in wavelet domain, EZW [1] and SPIHT [2] implicitly classify wavelet pixels as zerotree pixels or not, while MRWD[3], SLCCA[4], and EBCOT[5] implicitly classify wavelet pixels as neighbors of significant pixels or not. In this paper, the wavelet pixels to be encoded are explicitly and finely classified based on their predicted probabilities, which is more sophisticated and effective. Furthermore, wavelet pixel sorting is introduced to help improve rate-distortion performance within each bit-plane coding. The technique of pixel classification and sorting is simple, yet effective to produce the image code with excellent compression performance. In addition, our algorithm provides both SNR and resolution scalability.

Satellite images comprise very complex spatial features that challenge the traditional image processing and classification techniques. In the last decade, fractal compression of satellite images had emerged to ease and improve the classification strength of the machine learning classifiers. Image compression has become the most essential step to deal with the high-dimensional data produced by remote sensing applications. The paper represents the survey study focussing on analyzing the effect of fractal features on satellite image classification using various machine learning techniques. In the process, the recently published research papers, conference articles, authenticated writings published in Google Scholar, Elsevier, Springer, and other websites have been considered. The work also summarised the generalized steps followed by various literature cited satellite image classification works. The study also summarises the best classifier among the group of classifiers used by researchers to guide future researchers. The comparative analysis based on the class predicted using different classifiers had shown that with time fractal-based satellite image classification had gained much popularity over the past decade.

Feature selection and weighting are the primary activity of every learning algorithm for text classification. Traditionally these tasks are carried out individually in two distinct phases: the first is the global feature selection during a corpus pre-processing and the second is the application of the feature weighting model. This means that two (or several) different techniques are used to optimize the performances even if a single algorithm may have more chances to operate the right choices. When the complete feature set is available, the classifier learning algorithm can better relate to the suitable representation level the different complex features like linguistic ones (e.g. syntactic categories associated to words in the training materialor terminological expressions). In [3] it has been suggested that classifiers based on generalized Rocchio formula can be used to weight features in category profiles in order to exploit the selectivity of linguistic information techniques in text classification. In this paper, a systematic study aimed to understand the role of Rocchio formula in selection and weighting of linguistic features will be described.

This research is concerned with the application of neural network techniques to the problems of classifying images in a manner that is invariant to changes in position and scale. In addition to the goal of invariant classification, the network has to classify the objects in a hierarchical manner, in which complex features are constructed from simpler features, and use unsupervised learning. The resultant hierarchical structure should be able to classify the image by having an internal representation that models the structure of the image. After finding existing neural network techniques unsuitable, a new type of neural network was developed that differed from the conventional multi-layer perceptron type of architecture. This network was constructed from neurons that were grouped into feature detectors.These neurons were taught in an unsupervised manner that used a technique based on Kohonen learning.A number of novel techniques were developed to improve the learning and classification performance of the network. The network was able to retain the spatial relationship of the classified features; this inherent property resulted in the capability for position and scale invariant classification. As a consequence, an additional invariance filter was not required. In addition to achieving the invariance property, the developed techniques enabled multiple objects in an image to be classified. When the network had learned the spatial relationships between the lower level features, names could be assigned to the identified features. As part of the classification process, th e system was able to identify the positions of the classified features in all layers of the network. A software model of an artificial retina was used to test the grey scale classification performance of the network and to assess the response of the retina to changes in brightness. Like the Neocognitron, the resulting network was developed solely for image classification. Although the Neocognitron is not designed for scale or position invariance, it was chosen for comparison purposes because it has structural similarities and the ability to accommodates light changes in the image. This type of network could be used as the basis for a 2D-scene analysis neural network, in which the inherent parallelism of the neural network would provide simultaneous classification of the objects in the image.

Decision-making accompanies us on a daily basis, making it an integral part of life. Decisions are made on a wide range of subjects, some of which have a greater or lesser impact on other aspects of life. Decision-making also takes place in sport, where the aim is to strive for the best possible results from athletes. This paper uses the previously proposed model built based on the COMET method to evaluate swimmers' predispositions to compete in selected swimming styles. Additionally, we also examined the group of swimmers for versatility, and the evaluation was based on many criteria describing the physical parameters of the swimmers. The study showed that some athletes performed better when tested for versatility than for a specific swimming style, indicating that the athletes were misdirected. Therefore, the proposed model is a feasible approach to address proper styles assignment and maximising swimmers' performance.

최근 인터넷, IPTV/SMART TV, 소셜 네트워크 (social network)와 같은 정보 유통 채널의 다양화로 유해 비디오 분류 및 차단 기술 연구에 대한 요구가 높아가고 있으나, 현재까지는 비디오에 대한 유해성을 판단하는 연구는 부족한 실정이다. 기존 유해 이미지 분류 연구에서는 이미지에서의 피부 영역의 비율이나 Bag of Visual Words (BoVW)와 같은 공간적 특징들 (spatial features)을 이용하고 있다. 그러나, 비디오에서는 공간적 특징 이외에도 모션 반복성 특징이나 시간적 상관성 (temporal correlation)과 같은 시간적 특징들 (temporal features)을 추가적으로 이용하여 유해성을 판단할 수 있다. 기존의 유해 비디오 분류 연구에서는 공간적 특징과 시간적 특징들에서 하나의 특징만을 사용하거나 두 개의 특징들을 단순히 결정 단계에서 데이터 융합하여 사용하고 있다. 일반적으로 결정 단계 데이터 융합 방법은 특징 단계 데이터 융합 방법보다 높은 성능을 가지지 못한다. 본 논문에서는 기존의 유해 비디오 분류 연구에서 사용되고 있는 공간적 특징과 시간적 특징들을 특징 단계 융합 방법을 이용하여 융합하여 유해 비디오를 분류하는 방법을 제안한다. 실험에서는 사용되는 특징이 늘어남에 따른 분류 성능 변화와 데이터 융합 방법의 변화에 따른 분류 성능 변화를 보였다. 공간적 특징만을 이용하였을 때에는 92.25%의 유해 비디오 분류 성능을 보이는데 반해, 모션 반복성 특징을 이용하고 특징 단계 데이터 융합 방법을 이용하게 되면 96%의 향상된 분류 성능을 보였다. Recently, malicious video classification and filtering techniques are of practical interest as ones can easily access to malicious multimedia contents through the Internet, IPTV, online social network, and etc. Considerable research efforts have been made to developing malicious video classification and filtering systems. However, the malicious video classification and filtering is not still being from mature in terms of reliable classification/filtering performance. In particular, the most of conventional approaches have been limited to using only the spatial features (such as a ratio of skin regions and bag of visual words) for the purpose of malicious image classification. Hence, previous approaches have been restricted to achieving acceptable classification and filtering performance. In order to overcome the aforementioned limitation, we propose new malicious video classification framework that takes advantage of using both the spatial and temporal features that are readily extracted from a sequence of video frames. In particular, we develop the effective temporal features based on the motion periodicity feature and temporal correlation. In addition, to exploit the best data fusion approach aiming to combine the spatial and temporal features, the representative data fusion approaches are applied to the proposed framework. To demonstrate the effectiveness of our method, we collect 200 sexual intercourse videos and 200 non-sexual intercourse videos. Experimental results show that the proposed method increases 3.75% (from 92.25% to 96%) for classification of sexual intercourse video in terms of accuracy. Further, based on our experimental results, feature-level fusion approach (for fusing spatial and temporal features) is found to achieve the best classification accuracy.

In the real estate market, spatial features play a crucial role in determining property appraisals and prices. When spatial features are considered, classification techniques have been rarely studied compared to regression, which is commonly used for price prediction. This study reviews spatial features' effects on predicting the house price ranges for real estate in Istanbul, Turkey, in the classification context. Spatial features are generated and extracted by geocoding the address information from the original data set. This geocoding and feature extraction is another challenge in this research. The experiments compare the performance of Decision Trees (DT), Random Forests (RF), and Logistic Regression (LR) classifier models on the data set with and without spatial features. The prediction models are evaluated based on classification metrics such as accuracy, precision, recall, and F1-Score. We additionally examine the ROC curve of each classifier. The test results show that the RF model outperforms the DT and LR models. It is observed that spatial features, when incorporated with non-spatial features, significantly improve the prediction performance of the models for the house price ranges. It is considered that the results can contribute to making decisions more accurately for the appraisal in the real estate industry.

This paper aims to highlight how to use the computerized video analysis for learning the sports technique key elements of the start in swimming in the case of the students in higher education of other profiles. This scientific approach entailed the organization of an ascertaining experimental study, using the following research methods: bibliographic study of the specialized literature, video computerized method by means of Pinnacle Studio, Kinovea and Physics ToolKit programs, method of movement postural orientation, statistical-mathematical method which uses the KyPlot program and the method of graphical representation of results. The research was conducted from October 2017 to January 2018 and included 12 students (second-year - series no. 2) of the Faculty of General Medicine. The results of the study show the muscle strength development of arms, abdomen and legs of the students under research. The use of the video computerized method highlights and identifies the kinematic and dynamic characteristics of sports technique key elements used in swimming start regarding the launching posture, multiplication of body posture (flying through the air) and concluding body posture (entering the water). The propulsion and going out of water will be studied depending on the specific swimming style chosen by the students. The effective use of the computerized video analysis that deepens the understanding of sports technique phasic structure will allow the processing of the modern didactical programs for swimming learning. The modern research methods used in the video computerized analysis of sports technique of the start in swimming are an important help for the measurement, analysis and evaluation of the kinematic and dynamic structure of all swimming styles.