SEFPN: Scale-Equalizing Feature Pyramid Network for Object Detection

Feature Pyramid Network (FPN) is used as the neck of current popular object detection networks. Research has shown that the structure of FPN has some defects. In addition to the loss of information caused by the reduction of the channel number, the features scale of different levels are also different, and the corresponding information at different abstract levels are also different, resulting in a semantic gap between each level. We call the semantic gap level imbalance. Correlation convolution is a way to alleviate the imbalance between adjacent layers; however, how to alleviate imbalance between all levels is another problem. In this article, we propose a new simple but effective network structure called Scale-Equalizing Feature Pyramid Network (SEFPN), which generates multiple features of different scales by iteratively fusing the features of each level. SEFPN improves the overall performance of the network by balancing the semantic representation of each layer of features. The experimental results on the MS-COCO2017 dataset show that the integration of SEFPN as a standalone module into the one-stage network can further improve the performance of the detector, by ∼1AP, and improve the detection performance of Faster R-CNN, a typical two-stage network, especially for large object detection APL∼2AP.

Unusual Case of Low-Velocity Large Object Penetrating the Frontal Bone with Favorable Outcome: A Case Report in Pediatric Patient

BACKGROUND: Penetrating head injury is one of the deadliest forms of head trauma; the outcome is usually low, and patients who survive long enough require complex medical treatment. Immediate imaging assessment with appropriate management can improve patient recovery. CASE REPORT: We reported a case of head penetration trauma on a 12-year-old child with embedded cylindrical iron that enters the skull without any signs of neurological deficit. CONCLUSIONS: This case shows that aggressive diagnostic imaging and emergency care followed by proper immediate head surgery management and postoperative intensive care to monitor and intervene in possible surgical and medical complications could significantly improve patient outcomes.

Does the Size-Arrival Effect Occur With an Active Collision-Avoidance Task in an Immersive 3D Virtual Reality Environment?

Objective Determine whether the size-arrival effect (SAE) occurs with immersive, 3D visual experiences and active collision-avoidance responses. Background When a small near object and a large far object approach the observer at the same speeds, the large object appears to arrive before the small object, known as the size-arrival effect (SAE), which may contribute to crashes between motorcycles and cars. Prior studies of the SAE were limited because they used two dimensional displays and asked participants to make passive judgments. Method Participants viewed approaching objects using a virtual reality (VR) headset. In an active task, participants ducked before the object hit them. In a passive prediction-motion (PM) judgment, the approaching object disappeared, and participants pressed a button when they thought the object would hit them. In a passive relative TTC judgment, participants reported which of two approaching objects would reach them first. Results The SAE occurred with the PM and relative TTC tasks but not with the ducking task. The SAE can occur in immersive 3D environments but is limited by the nature of the task and display. Application Certain traffic situations may be more prone to the SAE and have higher risk for collisions. For example, in left-turn scenarios (e.g., see Levulis, 2018), drivers make passive judgments when oncoming vehicles are far and optical expansion is slow, and binocular disparity putatively is ineffective. Collision-avoidance warning systems may be needed more in such scenarios than when vehicles are near and drivers’ judgments of TTC may be more accurate (DeLucia, 2008).

Image processing of periapical radiograph on granuloma detection by analysis method based on Android

Objectives: The study assesses periapical radiograph image with various android based analysis method to detect granuloma. Materials and Methods: The study uses survey descriptive cross sectional by using questionnaire. The questionnaire is distributed to 70 random respondents. The methods of the android application used are BLOB (Binary Large Object), DCT and LDA (Discrete Cosine Transform and Linier Discriminant Analysis), DWT and PCA (Discrete Wavelet Transform & Principal Component Analysis), and multiwavelet transformation. The questionnaire assessment included accuracy, effectiveness, attractiveness, innovativeness of the android application. Results: Android application with BLOB has effectivity and accuracy of 62,5%, attractiveness and innovativeness of 75%. Android application with DCT and LDA has effectivity and accuracy of 50 %, attractiveness of 70% and innovativeness of 80%. Android application with DWT and PCA has effectivity of 50%, accuracy of 60%, attractiveness of 66,66% and innovativeness of 80%. Android application with multiwavelet transformation has effectivity and accuracy of 50%, attractiveness of 55% and innovativeness of 73%. Conclusion: Based on assessment, the four methods used to detect granuloma are effective and applicative with android-based application. Android-based Application can detect granuloma with approximately more than 70% successful rate. These methods ease the practitioner to interpret the granuloma image.

MINIMIZING THE EFFECTIVE AREA OF SCATTERING BY MEANS OF CHANGING THE RELATIVE MAGNETIC PERMEABILITY

Исследуется зависимость эффективной площади рассеяния (ЭПР) от относительной магнитной проницаемости материала, из которого изготавливается структура. В качестве тела моделирования был выбран шар, который изготовлен из диэлектрического материала, у которого возможно выполнять изменение относительной магнитной проницаемости. По полученным результатам моделирования построены графики зависимости максимального значения моностатической ЭПР от частоты, а также от относительной магнитной проницаемости среды. Было показано, что с увеличением относительной магнитной проницаемости материала изготовления происходит увеличение значения ЭПР объекта, а также обнаружена зависимость эффективной площади рассеяния от соотношения размеров шара и длиной волны, так при превышении порогового значения, после которого шар становится крупным объектом, ЭПР резко возрастает. По результатам исследования был построен график зависимости эффективной площади рассеяния шара от относительной магнитной проницаемости материала изготовления. Доказана возможность применения материала с частотозависимой относительной магнитной проницаемостью в качестве стелс-покрытия. В статье содержится исследуемая модель, графики полученных результатов, по которым можно легко определить зависимость ЭПР от частоты и от относительной магнитной проницаемости материала изготовления The article investigates the dependence of the effective scattering area (ESA) on the relative magnetic permeability of the material from which the structure is made. We chose a sphere as the modeling body, which is made of a dielectric material, in which it is possible to change the relative magnetic permeability. Based on the obtained simulation results, graphs of the dependence of the maximum value of monostatic ESA on frequency, as well as on the relative magnetic permeability of the medium, were constructed. It was shown that with an increase in the relative magnetic permeability of the material of manufacture, an increase in the value of the ESA of the object occurs, and the dependence of the effective scattering area on the ratio of the size of the ball and the wavelength was found, so when the threshold value is exceeded, after which the ball becomes a large object, ESA rises sharply. Based on the results of the study, a graph of the dependence of the effective scattering area of the sphere on the relative magnetic permeability of the material of manufacture was built. The possibility of using a material with a frequency-dependent relative magnetic permeability as a stealth coating was proven. The article contains the investigated model, graphs of the results obtained, by which it is easy to determine the dependence of the ESA on the frequency and on the relative magnetic permeability of the material of manufacture

ABOships—An Inshore and Offshore Maritime Vessel Detection Dataset with Precise Annotations

Availability of domain-specific datasets is an essential problem in object detection. Datasets of inshore and offshore maritime vessels are no exception, with a limited number of studies addressing maritime vessel detection on such datasets. For that reason, we collected a dataset consisting of images of maritime vessels taking into account different factors: background variation, atmospheric conditions, illumination, visible proportion, occlusion and scale variation. Vessel instances (including nine types of vessels), seamarks and miscellaneous floaters were precisely annotated: we employed a first round of labelling and we subsequently used the CSRT tracker to trace inconsistencies and relabel inadequate label instances. Moreover, we evaluated the out-of-the-box performance of four prevalent object detection algorithms (Faster R-CNN, R-FCN, SSD and EfficientDet). The algorithms were previously trained on the Microsoft COCO dataset. We compared their accuracy based on feature extractor and object size. Our experiments showed that Faster R-CNN with Inception-Resnet v2 outperforms the other algorithms, except in the large object category where EfficientDet surpasses the latter.

Sistem Identifikasi Usia Manusia pada Citra Panoramic Radiograph Gigi Molar Pertama

ABSTRAKOdontologi forensik merupakan sebuah cabang ilmu forensik yang melakukan proses identifikasi berdasarkan gigi. Gigi merupakan salah satu bagian tubuh manusia paling kuat kuat. Dalam masa pertumbuhan, gigi manusia mengalami degeneratif pada usia tertentu, sehingga gigi dapat menjadi media dalam proses identifikasi usia. Pada penelitian ini, dirancang sistem pengolahan citra yang dapat mendeteksi usia manusia pada citra radiograf panoramik gigi. Sistem ini menggunakan metode Binary Large Object dan Decision Tree. Berdasarkan hasil pengujian, sistem dapat mendeteksi usia berdasarkan citra gigi molar pertama dengan tingkat akurasi lebih dari 80%, pada saat menggunakan parameter structuring element jenis Disk dengan jari-jari 4 piksel, ciri area dan rasio pulpa, serta jenis algoritma pada decision tree yaitu curvature dengan jumlah 50 percabangan.Kata kunci: citra radiograf panoramik, pulpa gigi, molar pertama, decision tree, binary large object ABSTRACTForensic odontology is a branch of forensic science that carries out dental identification processes. Teeth are one of the strongest parts of the human body In the period of growth, human teeth degenerative at a certain age, so that teeth can be a medium in the process of age identification. In this study, an image processing system was designed that could detect human age on dental panoramic radiographs. This system using the Binary Large Object and Decision Tree methods. Based on the test results, the system can detect age based on the image of the first molar with an accuracy level of more than 80%, when using a Disk type structuring element parameter with a radius of 4 pixels, the area and pulp ratio features, and the type of algorithm in the decision tree, namely curvature with the number of 50 branches.Keywords: panoramic radiograph image, teeth pulp, first molar, decision tree,binary large object

Microbial Brazil nut effect

The Brazil nut effect (BNE) is a counter-intuitive process of segregation of a large object inside a vibrated granular medium (GM), which has been studied widely by subjecting GMs to...