Neural Network Analysis for Microplastic Segmentation

It is necessary to locate microplastic particles mixed with beach sand to be able to separate them. This paper illustrates a kernel weight histogram-based analytical process to determine an appropriate neural network to perform tiny object segmentation on photos of sand with a few microplastic particles. U-net and MultiResUNet are explored as target networks. However, based on our observation of kernel weight histograms, visualized using TensorBoard, the initial encoder stages of U-net and MultiResUNet are useful for capturing small features, whereas the later encoder stages are not useful for capturing small features. Therefore, we derived reduced versions of U-net and MultiResUNet, such as Half U-net, Half MultiResUNet, and Quarter MultiResUNet. From the experiment, we observed that Half MultiResUNet displayed the best average recall-weighted F1 score (40%) and recall-weighted mIoU (26%) and Quarter MultiResUNet the second best average recall-weighted F1 score and recall-weighted mIoU for our microplastic dataset. They also require 1/5 or less floating point operations and 1/50 or a smaller number of parameters over U-net and MultiResUNet.

Pemeringkatan Pencarian pada Buku Pedoman Akademik Filkom UB Menuju Merdeka Belajar dan Free E-Book Pembelajaran Sebagai Prototype Local Smart Micro Search Engine Menggunakan Algoritma Pagerank dan TF-IDF

<p class="Abstrak">Buku Pedoman Akademik FILKOM Universitas Brawijaya merupakan suatu kebutuhan informasi akademik yang cukup penting, dan juga buku penunjang pembelajaran seperti Free e-Book bagi para mahasiswa. Untuk memperoleh informasi yang relevan terhadap query yang diberikan seringkali belum sesuai dengan kebutuhan pencarian pengguna. Pengguna harus menguasai secara keseluruhan untuk mengetahui dokumen mana yang paling sesuai, dan proses ini akan memakan waktu yang banyak. Sistem ini mampu memberikan rekomendasi dokumen sesuai dengan hasil perhitungan pemeringkatan teks. Proses pemeringkatan teks dapat diselesaikan dengan algoritma PageRank, di mana dokumen yang memiliki bobot pemeringkatan terkecil, memiliki kata terbanyak pada dokumen tersebut. Algoritma ini telah dibuktikan mampu memeberikan feedback dokumen yang relevan melalui dua tahap pengujian. Evaluasi yang dilakukan terhadap dua buah pengujian menghasilkan rata-rata nilai recall tertinggi yaitu 80.6% pada data ke-1, dan data ke-2 didapatkan korelasi terbaik antara precision, recall dan f-measure sebesar 0,98, 0,99, 0,99.</p><p class="Abstrak"> </p><p class="Abstrak"><em><strong>Abstract</strong></em></p><p class="Abstract"><em>The Brawijaya University FILKOM Academic Handbook is an important academic information need, as well as learning support books such as Free e-Books for students. To obtain information that is relevant to the query given is often not in accordance with the wishes of the user. Users must master the whole to find out which documents are most suitable, which is where the process will take a lot of time. This system is able to provide document recommendations in accordance with the results of the text ranking calculation. The process of ranking the text can be solved by the PageRank algorithm, where documents that have the smallest ranking weight, have the most words in the document. This algorithm has been proven to be able to provide feedback on relevant documents through two stages of testing. he evaluation conducted on the two tests resulted in the highest average recall value of 80.6% on the 1st dataset, and 2nd dataset the best correlation was obtained between precision, recall and f-measure of 0.98, 0.99, 0.99.</em></p><p class="Abstrak"><em><strong><br /></strong></em></p>

A Deep Audiovisual Approach for Human Confidence Classification

Research on self-efficacy and confidence has spread across several subfields of psychology and neuroscience. The role of one’s confidence is very crucial in the formation of attitude and communication skills. The importance of differentiating the levels of confidence is quite visible in this domain. With the recent advances in extracting behavioral insight from a signal in multiple applications, detecting confidence is found to have great importance. One such prominent application is detecting confidence in interview conversations. We have collected an audiovisual data set of interview conversations with 34 candidates. Every response (from each of the candidate) of this data set is labeled with three levels of confidence: high, medium, and low. Furthermore, we have also developed algorithms to efficiently compute such behavioral confidence from speech and video. A deep learning architecture is proposed for detecting confidence levels (high, medium, and low) from an audiovisual clip recorded during an interview. The achieved unweighted average recall (UAR) reaches 85.9% on audio data and 73.6% on video data captured from an interview session.

Knowledge Graph-Based Image Recognition Transfer Learning Method for On-Orbit Service Manipulation

Visual perception provides state information of current manipulation scene for control system, which plays an important role in on-orbit service manipulation. With the development of deep learning, deep convolutional neural networks (CNNs) have achieved many successful applications in the field of visual perception. Deep CNNs are only effective for the application condition containing a large number of training data with the same distribution as the test data; however, real space images are difficult to obtain during large-scale training. Therefore, deep CNNs can not be directly adopted for image recognition in the task of on-orbit service manipulation. In order to solve the problem of few-shot learning mentioned above, this paper proposes a knowledge graph-based image recognition transfer learning method (KGTL), which learns from training dataset containing dense source domain data and sparse target domain data, and can be transferred to the test dataset containing large number of data collected from target domain. The average recognition precision of the proposed method is 80.5%, and the average recall is 83.5%, which is higher than that of ResNet50-FC; the average precision is 60.2%, and the average recall is 67.5%. The proposed method significantly improves the training efficiency of the network and the generalization performance of the model.

On the Limitations of Visual-Semantic Embedding Networks for Image-to-Text Information Retrieval

Visual-semantic embedding (VSE) networks create joint image–text representations to map images and texts in a shared embedding space to enable various information retrieval-related tasks, such as image–text retrieval, image captioning, and visual question answering. The most recent state-of-the-art VSE-based networks are: VSE++, SCAN, VSRN, and UNITER. This study evaluates the performance of those VSE networks for the task of image-to-text retrieval and identifies and analyses their strengths and limitations to guide future research on the topic. The experimental results on Flickr30K revealed that the pre-trained network, UNITER, achieved 61.5% on average Recall@5 for the task of retrieving all relevant descriptions. The traditional networks, VSRN, SCAN, and VSE++, achieved 50.3%, 47.1%, and 29.4% on average Recall@5, respectively, for the same task. An additional analysis was performed on image–text pairs from the top 25 worst-performing classes using a subset of the Flickr30K-based dataset to identify the limitations of the performance of the best-performing models, VSRN and UNITER. These limitations are discussed from the perspective of image scenes, image objects, image semantics, and basic functions of neural networks. This paper discusses the strengths and limitations of VSE networks to guide further research into the topic of using VSE networks for cross-modal information retrieval tasks.

IntSplice2: Prediction of the Splicing Effects of Intronic Single-Nucleotide Variants Using LightGBM Modeling

Prediction of the effect of a single-nucleotide variant (SNV) in an intronic region on aberrant pre-mRNA splicing is challenging except for an SNV affecting the canonical GU/AG splice sites (ss). To predict pathogenicity of SNVs at intronic positions −50 (Int-50) to −3 (Int-3) close to the 3’ ss, we developed light gradient boosting machine (LightGBM)-based IntSplice2 models using pathogenic SNVs in the human gene mutation database (HGMD) and ClinVar and common SNVs in dbSNP with 0.01 ≤ minor allelic frequency (MAF) &lt; 0.50. The LightGBM models were generated using features representing splicing cis-elements. The average recall/sensitivity and specificity of IntSplice2 by fivefold cross-validation (CV) of the training dataset were 0.764 and 0.884, respectively. The recall/sensitivity of IntSplice2 was lower than the average recall/sensitivity of 0.800 of IntSplice that we previously made with support vector machine (SVM) modeling for the same intronic positions. In contrast, the specificity of IntSplice2 was higher than the average specificity of 0.849 of IntSplice. For benchmarking (BM) of IntSplice2 with IntSplice, we made a test dataset that was not used to train IntSplice. After excluding the test dataset from the training dataset, we generated IntSplice2-BM and compared it with IntSplice using the test dataset. IntSplice2-BM was superior to IntSplice in all of the seven statistical measures of accuracy, precision, recall/sensitivity, specificity, F1 score, negative predictive value (NPV), and matthews correlation coefficient (MCC). We made the IntSplice2 web service at https://www.med.nagoya-u.ac.jp/neurogenetics/IntSplice2.

ANALISIS SISTEM PENATAAN DAN TEMU BALIK ARSIP INAKTIF DI DINAS PANGAN, TANAMAN PANGAN DAN HORTIKULTURA PROVINSI RIAU

This study aims to determine how the system of structuring and retrieval of inactive archives in the Department of Food, Crops and Horticulture Riau Province. The method used in this study is descriptive data analysis method. The data analysis techniques carried out by researchers are data reduction, data presentation and conclusion drawing. Meanwhile, archive retrieval refers to the calculation of recall-precision. The research population is all staff of the General and Civil Service Sub-Division who opened 25 people and inactive archives in 2015 which opened 327 archives. The research sample is 1 honorary staff of the general and civil service sub-section and 2 supervisory archivists and young expert archivists, researchers took 3 informants who were directly related to the arrangement of inactive archives. The archive sample used for archive retrieval is by using 8 keywords with archive classification codes, namely 005 (Invitation), 048 (Data Management), 822 (Regular Salary Increase), 823 (Raise), 842 (Funds), 851 ( Annual Leave), 855 (Leave for Hajj/Umrah), 862 (Penalty). The results showed that the inactive archive arrangement system at the Food, Food Crops and Horticulture Department of Riau Province was using guidelines based on the Regulation of the Head of the National Archives of the Republic of Indonesia Number 4 of 2017 and inactive archive retrieval using the Ms. The average recall value is 97.75% and the average recall level is 99.05%. So, the arrangement of inactive archives is quite good and to remember the precision of the archive shows the results of an unbalanced proportion due to several factors, namely there are still archives that have not been inputted and there are still archives that do not match the keywords used.

Applying Similarity Measures to Improve Query Expansion

The huge evolving in the information technologies, especially in the few last decades, has produced an increase in the volume of data on the World Wide Web, which is still growing significantly. Retrieving the relevant information on the Internet or any data source with a query created by a few words has become a big challenge. To override this, query expansion (QE) has an important function in improving the information retrieval (IR), where the original query of user is recreated to a new query by appending new related terms with the same importance. One of the problems of query expansion is the choosing of suitable terms. This problem leads to another challenge of how to retrieve the important documents with high precision, high recall, and high F measure. In this paper, we solve this problem through applying different similarity measures with the use of English WordNet. The obtained results proved that, with a suitable selection method, we are able to take advantage of English WordNet to improve the retrieval efficiency. The work proposed in this paper is extracting the terms from all the documents and query, then applying the following steps: preprocessing, expanding the query based on English WordNet, selecting the best terms, weighting of term, and finally using the cosine similarity and Jaccard similarity to obtain the relevant documents. Our practical results were applied on the DUC2002 dataset that contains 559 documents distributed over several categories. The average precision of cosine (for random queries) = 100% whereas the average precision of Jaccard = 84.4 %, and the average recall of cosine = 86.8% whereas the average recall of Jaccard = 73.4%. The average f-measure of cosine = 92%, whereas the average f-measure of Jaccard = 76%.

Klasifikasi Citra Histopatologi Kanker Payudara menggunakan Data Resampling Random dan Residual Network

Data imbalance between classes is one of the problems on image classification. The data in each class is not equal and has a relatively large difference generated in less than optimal classification results. Ideally, the data in each class is equal or have a slight difference. This article discusses the classification of the histopathology breast cancer image. The data consist of 8 classes with unbalanced data. The method for balancing the data in each class uses random resampling which is applied to training data only. The data used from BreakHist with magnifications of 40x, 100x, 200x, and 400x . The classification uses Residual Network (ResNet) 18 and 50. The best results are obtained on images with a magnification of 400x. Classification results using ResNet18 has an average accuracy of 79.82%, an average precision of 71.39%, and an average recall of 82.37%. Meanwhile using ResNet50 showed an average accuracy of 81.67%, average precision of 78.41%, and an average recall of 82.91%.

Towards Automatic Modeling of Volleyball Players’ Behavior for Analysis, Feedback, and Hybrid Training

Automatic tagging of video recordings of sports matches and training sessions can be helpful to coaches and players and provide access to structured data at a scale that would be unfeasible if one were to rely on manual tagging. Recognition of different actions forms an essential part of sports video tagging. In this paper, the authors employ machine learning techniques to automatically recognize specific types of volleyball actions (i.e., underhand serve, overhead pass, serve, forearm pass, one hand pass, smash, and block which are manually annotated) during matches and training sessions (uncontrolled, in the wild data) based on motion data captured by inertial measurement unit sensors strapped on the wrists of eight female volleyball players. Analysis of the results suggests that all sensors in the inertial measurement unit (i.e., magnetometer, accelerometer, barometer, and gyroscope) contribute unique information in the classification of volleyball actions types. The authors demonstrate that while the accelerometer feature set provides better results than other sensors, overall (i.e., gyroscope, magnetometer, and barometer) feature fusion of the accelerometer, magnetometer, and gyroscope provides the bests results (unweighted average recall = 67.87%, unweighted average precision = 68.68%, and κ = .727), well above the chance level of 14.28%. Interestingly, it is also demonstrated that the dominant hand (unweighted average recall = 61.45%, unweighted average precision = 65.41%, and κ = .652) provides better results than the nondominant (unweighted average recall = 45.56%, unweighted average precision = 55.45, and κ = .553) hand. Apart from machine learning models, this paper also discusses a modular architecture for a system to automatically supplement video recording by detecting events of interests in volleyball matches and training sessions and to provide tailored and interactive multimodal feedback by utilizing an HTML5/JavaScript application. A proof of concept prototype developed based on this architecture is also described.