Search

Diffuse optical tomography (DOT) is promising to investigate functional and structural information in biological tissue by involving the forward problem and inverse solution. The work within this study develops an image reconstruction algorithm based on frequency-domain DOT using multi-frequency. In the forward problem, numerical simulation performs a diffusion equation to compute the light radiance distribution inside the model geometry with the assumed light intensity and optical properties. While in inverse solution, the absorption and reduced scattering coefficients are reconstructed utilizing light boundary data. To test the algorithm, several simulation cases, such as different contrast ratios between background and inclusion, inclusion size, as well as single and multi-frequency using one to three frequencies each and simultaneously are implemented. For verifying the image reconstruction results of multi-frequency, an objective numerical assessment, namely contrast and size detail is applied with a threshold value to specify the visible inclusion inside the geometry. Reconstructed images indicate that the effectivity of two frequencies in multi-frequency with 100 and 150 MHz has similar results as in single-frequency with 200 MHz. Therefore, the more number of frequencies may not enhance the contrast and image quality.

This paper proposed a grape drying machine using computer vision and Multi-layer Perceptron (MLP) method. Computer vision is for taking grapes’ image on conveyor, whereas MLP is for controlling grape drying machine and classifying its output. To evaluate the proposed, a kind of grapes are put on conveyor of the machine and their images are taken every two min. Some parameters of MLP to control the drying machine includes dried grape, temperature, grape area, motor position, and motion speed. Those parameters are to adjust an appropriate MLP’s output, including motion control and heater control. Two different temperatures are employed on the machine, including 60 and 75°C. The results showed that the grape could be dried with similar area 3800 pixel at the 770th min using temperature 60°C and at the 410th min using temperature 75°C. Comparing between them, the similar ratio could also be achieved at 0.64 with different time 360 min. Indeed, the temperature setting at 75°C resulted faster drying performance.

A public database containing representative data of karawitan traditional music is needed as a resource for researchers who study computer music and karawitan. To establish this database, a text-based pitch model for music representation that is both human and computer-based was first investigated. A new model of musical representation that can be read by humans and computers is proposed to support music and computer research on karawitan also known as gamelan music. The model is expected to serve as the initial effort to establish a public database of karawitan music representation data. The proposed model was inspired by Helmholtz Notation and Scientific Pitch Notation and well-established, text-based pitch representation systems. The model was developed not only for pitch number, high or low or middle pitch information (octave information), but for musical elements found in gamelan sheet music pieces that include pitch value and legato signs. The model was named Gendhing Scientific Pitch Notation (GSPN). Ghending is a Javanese word that means “song”. The GSPN model was designed to represent music by formulating musical elements from a sheet music piece. Furthermore, the model can automatically be converted to other music representation formats. In the experiment, data in the GSPN format was implemented to automatically convert sheet music to a binary code with localist representation technique.

The performance appraisal of Sunday school teacher in the Imanuel Lurang congregation aims to measure and distinguish the quality of performance achieved by Sunday school teacher and decide various policies such as giving rewards to every Sunday school teacher with the best performance, and for Sunday school teacher who have poor performance scores will be given a guiding, approach, etc. The number of criteria in determining the quality of Sunday school teacher is not an easy thing to do by manual. Then it is essential that a computerized performance appraisal-based performance app can speed up the process of progressing to be more effective and efficient. This research develops decision support systems (DSS) that is dynamic using the PHP programming language, by combining the AHP method that has been refined by the VIKOR method. The AHP method is used in determining the weight of each criterion, and the VIKOR method is used for the ranking process. Test results indicate that the system can provide a sequence of alternative Sunday school teacher that will be used as recommendations for decision makers to determine which Sunday school teachers are quality and not qualified.

The class imbalance is a condition when one class has a higher percentage than the other then it can affect the accuracy. One method in data mining that can be used to classification is logistic regression method. The method used in this research is RWO-sampling method using random replicate approach for synthetic data generation on descrete attribute. The result of the research can handle the problem of class imbalance, RWO-sampling method with random replicate approach shows better accuracy than RWO-sampling method with roulette and ROS approach. The accuracy value for RWO-Sampling method with roulette and RWO-Sampling approach with random replicate approach has increased to an average of 15.55% of each dataset. As for comparithem with the ROS method has increased an average of 3.7% of each dataset. Furthermore, for testing the underfitting problem in logistic regression, the oversampling method is better than non-oversampling with an increase in accuracy value reaching an average of 2.3% of each dataset.

The farming profession in the industrial era 4.0 is considered ancient and has no potential and bonafide. To change such a mindset, Community Service (PKM-Unmer Building Village) in collaboration with hydroganic farmer groups to make greenhouse and hydro-gas models accompanied by water flow monitoring (flow sensor), temperature, and humidity monitoring (DHT11 sensor) and electricity control (UPS). By utilizing the IoT concept in a monitoring tool that is placed in hydroganics, it can monitor the flow of water and plants in real-time. To access monitoring, the cellphone must be connected to the same wifi as the device, then enter the IP address using a mobile browser. The method of implementation used consists of three stages: preparation includes a review of partner locations, coordination, and the number of partners and existing facilities, pilot and, socialization. The results obtained from this activity are to accelerate and make it easier to control the hydroganics contained in the greenhouse in real-time. By utilizing a UPS to make the water pump always on, the equipment is not easily damaged if there is an electrical disturbance such as a blackout or short circuit. PKM-Unmer Building Village produces a pilot modeled in a greenhouse, socialization to the community and Bengkel Mimpi farmer groups, and training to make simple monitoring tools.

In the construction of Cable Stayed Bridge, the use of heavy construction equipment is absolutely necessary, especially required for positioning and selecting the type of Tower Crane in order to improve construction performance. Therefore, human resources and heavy equipment used in the construction process will significantly affect the period of project implementation. Proper planning of the location for various construction equipment is believed to be the key of efficiency and productivity. Tower Crane as the target optimization is one of the equipment that has a big role in transporting the materials for the construction of a multi- storey building so that it requires proper planning to be able to serve all points of demand from its position.

Traffic is a medium to move from one point to another. Therefore, the role of traffic is very important to support vehicle mobility. If congestion occurs, mobility will be hampered so that it gives influence to other sectors such as financial, air pollution and traffic violations. This study aims to create a model to predict vehicle queue at the traffic lights when its status is red. The prediction is conducted by using Neural Network with Extreme Learning Machine method to predict the length of the vehicle queue, and Correlation Analysis was used to measure the correlation between the connected roads. The conducted experiments use data of the length of the vehicle queue at the traffic lights which was obtained from DISHUB (Transportation Bureau) DI Yogyakarta. Several experiments were carried out to determine the optimum prediction model of vehicle queue length. The experiments found that the optimum model had an average MAPE value of 15.5882% and an average Running Time of 5.2226 seconds.

The flow of containers in the terminals of the Port of Tanjung Perak Surabaya has increased. The increased container flow has an impact on the pilling up of ex-wood pallets and if this left continuously it become city trash. Even though ex-wood pallets have the characteristics and potential to be developed, for that we have do upcycling. In addition, people who live in narrow homes usually discarded or replaced their old interior products because they do not fit into the new space. Therefore, people need a modular interior product that can be assembled from scratch, and can be arranged according to the needs and desires of the user, by doing upcycling to reduce the pilling up of ex-wood pallets. The design aims to knowing the characteristics and potential of ex-wood pallets, then reused into modular interior products with modules that accordance with their characteristic. The modules are connected using connectors, which can be assembled according to the user’s needs. Design thinking methods consisting of inspiration, ideation, and implementasion are carried out in this design process. The results of this design are modules and connectors that can be assembled into 12 types of products, media promotion, and booth design. This design proved that ex-wood pallets have characteristics and potential that can be develop by upcycling into modules in accordance with the characteristics of used wood pallets that can be assembled using connector into a variety of modular interior products according to the needs and desires of the user.

Background: This study aims to develop and produce a diagnostic tool capable of detecting student misconceptions on the concept of ecosystems. Methods: This study used a research and development method with a modified 4D design (Define, Design, Develop and Disseminate) without disseminating or spreading. Data collection methods used in this research and development are observation, literature review, documentation, questionnaires, interviews and tests. Results: The results showed that the instruments developed were valid by validating 3 experts, namely material experts, evaluation and assessment experts, and linguists with an average percentage of each feasibility of 90% material expert judgment, 82% assessment expert assessment and evaluation. and 80% linguist assessment. Meanwhile, the instrument readability test showed a percentage of 63% in good category. The validity value of the instrument in general was 0.70 with the valid or high validity category, while the instrument reliability value was 0.82 with the very high category of reliability. The overall results of students' misconceptions on ecosystem material showed 23% of students experienced low-category miconceptions, 8% of students experienced no conceptual understanding, 17% of students experienced conceptual understanding, and 52% of students experienced partial understanding. Conclusion: Based on these results it can be said that the four-tier test instrument developed is feasible and can diagnose misconceptions in students

PT. XXX mendapatkan sampah-sampah plastik yang tidak terpakai dari supplier penyedia sampah-sampah dalam negeri maupun sampah-sampah dari luar negeri. Dalam proses pengolahan sampah plastik ini, PT. XXX masih menggunakan tenaga sortir secara manual yang berjumlah 40 orang, hasil yang diperoleh juga kurang maksimal karena masih banyak material lain yang tercampur dalam sampah plastik dari 4200 kg sampah plastik hanya dapat memisahkan material lain yang tidak terpakai sebanyak 915,92 kg per harinya. Penelitian ini bertujuan untuk merancang mesin Centrifugal sebagai alat pemisah antara sampah plastik yang terpakai dari material lain yang tidak terpakai serta meminimalisir tenaga sortir sebanyak 40 orang menjadi 8 orang, dan memaksimalkan pemisahan material yang tidak terpakai menjadi 1355,308 kg per hari. Dari perancangan mesin Centrifugal ini mampu mengatasi masalah yang ada pada bagian sortir sehingga dapat menghasilkan biji plastik yang lebih bersih dan mempercepat proses penyortiran.

The aims of the study were analyzing students’ understanding and skills to make concept mapping in work energy matter through C-Map application software. True experimental method with randomized control group pre-test and post-test design was utilized in the research method. The research involved 72 public high school students from second semester, 10th grade in Indonesia, composed of 36 students in an experiment group (EG) and 36 students in control group (CG). Data were collected by giving a test of work and energy to evaluate students’ understanding through C-Map application during Spring Semester 2019. A specific rubric from the previous researchers was used to evaluate students’ skills in developing concept map. Analysis of data used independent t-test and normalized gain (Ngain). The independent test revealed the implementation of C-Map application affected in improving students’ understanding of work and energy. Students’ understanding of these concepts performed that EG higher than CG. It was supported by the obtained of normalized gain (N-gain). The N-gain of EG and CG was 0.39 and 0.26, respectively. Students’ skill to make concept mapping in the topic of work and energy through C-Map application achieved 81.36 of 100 with a very good category. C-Map application guide students to arrange the hierarchical structure concept of work and energy, construct concept from focus question and see how the concept related with another concept.

Faces is a unique region in our body that can be used as a biometric identity. Furthermore, the face between two people that have a kinship relationship may share the same face features which can be used to decide whether two people have a kinship relationship or not. In this paper, we proposed a family-aware convolutional neural network (CNN) for the visual kinship verification problem. Our proposed classifier is constructed by paralleling the state-of-the-art face recognition model and attaching two additional networks, a family-aware network, and a kinship verification network. The family-aware network weights adjusted by learning features specific to the family using deep metric learning loss while the kinship verification network use softmax loss to learn the kinship verification problem. One of the advantages of our proposed classifier is that the output of the classifier is normalized and can be represented as the probability of two images being kin or non-kin. To preserve the face recognition features extraction ability in the state-of-the-art face recognition model, we freeze the weights of the convolutional layers in the classifier for the training process. In the testing process, the family-aware network is detached to construct the final classifier. Experiments on FIW (Families In the Wild) dataset show that our proposed classifier performs better comparing with classifiers that trained without a family-aware network and the ensemble version of the classifier is comparable with several state-of-the-art methods with an average accuracy of 68.84%.

Most wheelchairs circulating in Indonesian market are imported from China at affordable prices, but do not fufill the ISO 7176 series standards. SNI 09-4663-1998 about wheelchair is one of medical devices standard. The standard is considered do not covers the quality and user safety, while wheelchair technology is transform rapidly. Mapping and analyzing gap between stakeholder need and SNI 09-4663-1998 is expected to be input of standard review to improve the competitiveness of products. This research using the Zachman Framework approach following FACTS steps. Data collection uses the observation, in-depth interviews and Focus Group Discussion. Data analysis using Inter Rater Reliability (IRR) which to measure the level of agreement of respondents. The results showed that stakeholder technical needs included 7 parameters: stability, maneuverability, mobility, dimensions, strength, durability, and product information. SNI 09-4663-1998 only includes the product strength parameters through a drop test so that the standard needs to be reviewed.

The function of KUA in the activities surrounding the religion of Islam, including providing service and guidance in the area of present services in terms of marriage and reconcilemen t for Muslims, provide services and guidance in the field of development of Sakina, family consultation conflict or household problems, and so on. Integration between the computer and artificial intelligence into the post-wedding consulting services is one approach in overcoming the limitations of the expert (religious instructor). This research aims to identify conflict in marriage by applying Naive Bayes algorithm at the stage of determining the groups of test data (retrieve), then entered the stage of the search process of the highest similarity value by using the Nearest Neighbor algorithm (reuse). The data source and the test data used are divided into two groups, namely marriage, and history data consultation, While the group conflicts are identified will be divided into five classes, namely an employment factor, the factor of age, educational factors, factors the number of weddings, and social status. Testing is performed by the use of 12 data, consisting of 11 data cases and 1 test data. At the stage of determination of group conflict acquired test data included in group one i.e. F001 (factor of the job), so at the stage of looking for value similarities used only the base case of the class F001 i.e. KK001, KK003, and KK008. The KK001 similarity has a value of 0476, KK003 of 0882, and KK008 of 0142. The case with most high similarity value will be stored as a base case. If the value similarity obtained less than the threshold value that is 0.8, then the solution of the case will be revised by experts. The results of the calculation accuracy, using 35 new test data that gets the value of 82.86% .

This study is development research with the aim the knowing of validity, effectiveness, and practicality of inquiry-based learning tools assisted by flashcard media to reduce misconceptions of elementary school students on light concept. Before and after the learning materials implemented, the pretest and posttest were carried out. Identification of students' misconceptions was utilized by using the three-tiers diagnostic test. The data analysis technique used quantitative descriptive. The results of the study showed: (1) The validity of the learning materials was declared to be very valid by the validator, (2) The practicality of the materials developed was declared practical by the observer in terms of aspects of the device's implementation and student activities in learning which were categorized very well, and (3) the effectiveness of the learning materials being developed was declared effective in terms of the application of the materials which had an impact on reducing students' misconceptions which initially amounted to 57.33% down to 17.33%. An Inquiry-based science learning device assisted with flashcard media that has been developed, was feasible with valid, practical, and effective criteria, and also can be used to reduce science misconceptions on elementary students on light material.

The process of registering new students (PMB) for sharing information displayed on the website based PMB information system. However, constraints in the usage of the website are prospective students must monitor the website frequently to obtain information. If not accordingly a student is at risk of lagging behind the latest information. The research aims to expand a registration information system on the enrollment of student-based websites and developed into an Android-based application. So it is very easy for students to access PMB information, and registration through the Android application. This study designed an Android-based registration information system application using the waterfall method, brute force algorithm, integrated realtime firebase database and firebase storage cloud. The results of testing the brute force algorithm on the word search "S1 Informatics Engineering" with the input string "Informatics" carried out 8 steps of searching. These results by calculating the parameters of cyclomate complexity, region, and independent path from whitebox tests obtained the same results, the amount of 16. The results of testing blackbox method declared successful and as expected. So it can be concluded the development of an online smart register information apply the brute force algorithm is said to be feasible to implementation in the process of registering new students in accordance with the testing that has been done.

Human–object interaction (HOI) detection is a popular computer vision task that detects interactions between humans and objects. This task can be useful in many applications that require a deeper understanding of semantic scenes. Current HOI detection networks typically consist of a feature extractor followed by detection layers comprising small filters (eg, 1 × 1 or 3 × 3). Although small filters can capture local spatial features with a few parameters, they fail to capture larger context information relevant for recognizing interactions between humans and distant objects owing to their small receptive regions. Hence, we herein propose a three‐stream HOI detection network that employs a context convolution module (CCM) in each stream branch. The CCM can capture larger contexts from input feature maps by adopting combinations of large separable convolution layers and residual‐based convolution layers without increasing the number of parameters by using fewer large separable filters. We evaluate our HOI detection method using two benchmark datasets, V‐COCO and HICO‐DET, and demonstrate its state‐of‐the‐art performance.

Based on Article 10 paragraph 1 of Law No. 14 of 2005, a teacher must have four competencies: pedagogical, personality, social, and professional. ICT training at Sunan Kalijaga State Islamic University involves instructors as educators who must have such competencies. An instructor's performance is assessed through students' learning evaluation system by giving comments to the instructions. These comments contain positive and negative sentiments that can be reviewed by conducting sentiment analysis. Research related to sentiment analysis in recent years has been widely done, but researchers rarely pay attention to the effect of sentence length from the dataset on the method's performance. This study tried to analyze sentiment related to sentence length effect on ICT training student comments using Support Vector Machine and Convolutional Neural Network methods. This study concluded that the sentence length on the dataset would affect the SVM and CNN methods' performance when combined with Word2vec. While the SVM+TFIDF method performance is not affected by sentence length, this method has the fastest process time than other methods. The CNN+Word2vec method produced the best performance in this study with a value of 0.94% accuracy, 0.95% precision, 0.96% recall, and 0.95% f1-score.

Traditional Javanese compositions contain melodies and skeletal melodies. Skeletal melodies are an extraction form of melodies. The melody extraction problem is similar to the chord detection in Western music, where chords are extracted from a melody. This research aims to develop a melody extraction system for traditional Javanese compositions. Melodies which have a time series data structure were designed as a part of the supervised learning problem to be solved using the pattern recognition technique and the Feed-Forward Neural Networks method. The melody data source uses a symbolic format in the form of sheet music. The beats in melodies data are used as the input and notes in skeletal melodies are used as the target. An FFNN multi-class classifier was built with six classes as the targets, where the class represents notes of the musical scale system. The network evaluation was conducted using accuracy, precision, recall, specificity and F-1 score measurements.

Synchronization of dual servo motor control has become an important issue due to many applications in engineering fields, such as electric vehicle, robotics, electronics production machines, and others. This paper studies cerebellar model articulation controller (CMAC) neural network (NN) controller to synchronize two servo motors with dynamic LuGre friction model. CMAC is kind of NN method represented by associative memory with more powerful properties. Cross-coupling control structure is employed to synchronize two servo motors in this study. To investigates the performance, MATLAB Simulink is applied to simulate the control design of dual servo motor. The simulation results exhibit that CMAC controller has better output trajectory and works well for two servo motors with different parameters of LuGre friction model.

The rapid development of deep learning improves the detection and classification of attacks on intrusion detection systems. However, the unbalanced data issue increases the complexity of the architecture model. This study proposes a novel deep learning model to overcome the problem of classifying multi-class attacks. The deep learning model consists of two stages. The pre-tuning stage uses automatic feature extraction with a deep autoencoder. The second stage is fine-tuning using deep neural network classifiers with fully connected layers. To reduce imbalanced class data, the feature extraction was implemented using the deep autoencoder and improved focal loss function in the classifier. The model was evaluated using 3 loss functions, including cross-entropy, weighted cross-entropy, and focal losses. The results could correct the class imbalance in deep learning-based classifications. Attack classification was achieved using automatic extraction with the focal loss on the CSE-CIC-IDS2018 dataset is a high-quality classifier with 98.38% precision, 98.27% sensitivity, and 99.82% specificity.

The problem of digital image transmission lies on the size of the image which tends to get bigger, and how to satisfy the confidentiality aspects when images are transmitted via unsecured channels. To solve this problem, a cryptosystem model is proposed, using a combination of compression and cryptography methods. This research proposes the development and performance measurement of compression-encryption models, Which consists of Generative Adversarial Network and logistic map encryption. The test results also showed that this scheme is better than other color image compression-encryption algorithms, With the SSIM values above 0.97 and compression ratio above 94%.

Recently, nonlinear system identification has received increasingly more attention due to its promising applications in engineering fields. It has become a challenging task to truly apply this system due to many complex factors especially the nonlinear and dynamical properties. The objective of this paper is to design a nonlinear system identification method using an appropriate learning method. This paper proposes a biogeography-based optimization (BBO)-based multilayer perceptron (MLP) architecture for nonlinear system identification. The BBO algorithm with its habitats imitates the species migration between them. A good solution is featured by an island with a higher High Suitability Index (HSI), and a poor solution by an island with a lower HSI. Higher HSI solutions resist change more effectively than lower HSI solutions. By combining the two schemes, the proposed MLP architecture with BBO learning provides a promising scheme for nonlinear system identification. Three kinds of nonlinear system are adopted for experimental verification, including Mackey–Glass series, Henon system, and nonlinear plant system. Mean squared error (MSE) index is used to calculate the difference between the measured input and output of the systems. By employing the nonlinear cases, the proposed algorithm presents rapid convergence and excellent MSE in nonlinear system identification.

Enterprise Resource Planning (ERP) packages are important for companies to integrate previously partial departments into one integrated system. This integrated system can incorporate data from various departments to facilitate decision making that fits the companys internal and external circumstances. The implementation of ERP in companies needs some people or key users who are able to work on the ERP system properly and correctly. The capabilities of the key users are related to soft ERP and hard ERP. Human-related abilities are called soft ERP while the equipment and tools are called hard ERP. Together, the soft ERP and hard ERP define the data integration in a company and become one entity in the well-being of communication among departments. The data for this research was collected using the likert scale. From the 100 questionnaires distributed to companies, there are 90 returned questionnaires or 90% response rate. The data processing with Java Web Start software showed the result of positive significant impact of 0.643 from soft ERP to hard ERP. The Soft ERP has a positive and significant impact of 0.262 on integrated ERP. The Soft ERP has a direct influence of 0.281 on the company performance. The Hard ERP has a positive and significant impact of 0.439 on integrated ERP. The Hard ERP has a positive and significant impact of 0.375 on firm performance. The integrated ERP has a direct influence of 0.338 on company performance. Overall, the impact of the total variables of independent Soft ERP to company performance is equal to 0.706. Therefore, the integration of ERP system brings positive and significant impacts to the business performance of manufacturing companies in East Java, Indonesia

Detection of seizure from EEG signals may help neurologist analyze the condition of epileptic patients. Variance in types of epileptic seizures provides challenge in recognizing the pattern of seizure signals from normal ones. In this paper, we discuss the classification of seizure and non-seizure conditions in epilepsy based on bandwidth features of EEG signals. These features were extracted using Empirical Mode Decomposition (EMD). This method decomposes signal into eight Intrinsic Mode Functions (IMFs) and its residue. The IMFs is then analyzed by Hilbert transform to obtain amplitude modulation bandwidth (BAM) and frequency modulation bandwidth (BFM). These features are fed as input to Support Vector Machine (SVM). We propose to combine the first four IMFs to extract bandwidth features and we implement this method on two data sets, public data set that contains signals from both extracranial and intracranial EEG, and data set OF Rumah Sakit Universitas Airlangga (RSUA) that contains signals from extracranial EEG only. The results showed that in general, the values of accuracy and specificity using combined features of first four IMFs outperformed those using features of single IMF, for all SVM kernel functions. The best accuracy and specificity on public data set were 97.3% and 98.25% respectively using RBF kernel, while the best sensitivity was gained using polynomial kernel based on only IMF1 bandwidth features by 99.0%. The experiments on RSUA data showed that the best accuracy was 96.5% and the best specificity was 99.38% when using Mexican Hat kernel based on combined features of first four IMFs, while the best sensitivity was 94% when using polynomial kernel based on only IMF1 bandwidth features.

In this paper, we investigated dual VGG-Face convolutional neural network classifier for image-based kinship verification problems. The proposed classifier is formed by paralleling convolutional layers of VGG CNN architecture and combined it with several fully-connected layers. Although the VGG CNN architecture is consists of huge parameters, the number of parameters in the classifier will be reduced by more than 80% by removing the fully-connected layers of the original classifier. We use the multi-task loss function for the training process to ensure that the features learned by the classifier are also can be used for family classification. Experiments on the FIW kinship verification dataset show that dual VGG-Face CNN classifiers can achieve an average accuracy of 64.71% on a single classifier and 65.49% on ensemble configuration. Based on our experiments, the lowest accuracy is produced for second-generation kinship (grandchild - grandparent) which have a low number of examples on FIW kinship.

Rapid technological developments have penetrated the application of technology in the field of aquaculture. Climate Changes that are currently uncertain and beyond prediction encourage the application of technology to increase fishery production and modernization. The purpose of this study is to design a water quality monitoring system that can be monitored remotely in real-time using an internet-based smartphone/PC application. It is equipped with an early warning system to monitor the water quality set by the upper and lower thresholds. This device is based on Robotdyn Uno Wifi R3 ESP8266 with IoT technology. The alarm system uses a buzzer and LED indicator to make it understandable to fishermen. We tested the device in two stages. Results showed that the alarm system responds to all changes in water quality that exceeds the set threshold. Water quality data can be monitored with an internet-based smartphone application and updated every 15-16 seconds

One of the goals of implementing a smart campus is to improve public services. Whereas the Academic Information System (Siakad) is a very important type of management information system (MIS) digital service. To find out the quality of service from the viewpoint of the user of this information system, a usability test was made. Usability is a qualitative analysis in determining the quality of an information system based on the user. The purpose of this study is to conduct usability testing SIAKAD in IAIN Salatiga using the System Usability Scale (SUS) method. Based on the usability test results, the majority of respondents use and utilize this information system well, this is proven by the value of the test results of 84,75. The evaluation results show that the Academic Information System at IAIN Salatiga is categorized as excellent.

In daily life, the deaf use sign language to communicate with others. However, the non-deaf experience difficulties in understanding this communication. To overcome this, sign recognition via human-machine interaction can be utilized. In Indonesia, the deaf use a specific language, referred to as Indonesia Sign Language (BISINDO). However, only a few studies have examined this language. Thus, this study proposes a deep learning approach, namely, a new convolutional neural network (CNN) to recognize BISINDO. There are 26 letters and 10 numbers to be recognized. A total of 39,455 data points were obtained from 10 respondents by considering the lighting and perspective of the person: specifically, bright and dim lightning, and from first and second-person perspectives. The architecture of the proposed network consisted of four convolutional layers, three pooling layers, and three fully connected layers. This model was tested against two common CNNs models, AlexNet and VGG-16. The results indicated that the proposed network is superior to a modified VGG-16, with a loss of 0.0201. The proposed network also had smaller number of parameters compared to a modified AlexNet, thereby reducing the computation time. Further, the model was tested using testing data with an accuracy of 98.3%, precision of 98.3%, recall of 98.4%, and F1-score of 99.3%. The proposed model could recognize BISINDO in both dim and bright lighting, as well as the signs from the first-and second-person perspectives.

Nowadays, speech is used also for communication between humans and computers, which requires conversion from speech to text. Nevertheless, few studies have been performed on speech-to-text conversion in Indonesian language, and most studies on speech-to-text conversion were limited to the conversion of speech datasets with incomplete sentences. In this study, speech-to-text conversion of complete sentences in Indonesian language is performed using the deep bidirectional long short-term memory (LSTM) algorithm. Spectrograms and Mel frequency cepstral coefficients (MFCCs) were utilized as features of a total of 5000 speech data spoken by ten subjects (five males and five females). The results showed that the deep bidirectional LSTM algorithm successfully converted speech to text in Indonesian. The accuracy achieved by the MFCC features was higher than that achieved with the spectrograms; the MFCC obtained the best accuracy with a word error rate value of 0.2745% while the spectrograms were 2.0784%. Thus, MFCCs are more suitable than spectrograms as feature for speech-to-text conversion in Indonesian. The results of this study will help in the implementation of communication tools in Indonesian and other languages.

The human face can be used in various biometrics procedures to identify an individual through face recognition or for facial expression recognition. However, not many studies have addressed the problem of face recognition along with facial expression recognition. In addition, some studies have directed more attention to finding the most suitable feature to extract and feed to a classifier. This study focused on addressing the problem using a convolutional neural network (CNN)-based method. Unlike other methods that require suitable features to be found, this study utilized raw images as the input to the CNN. A total of 16,640 images showing four facial expressions (normal, smiling, surprised, and angry) were used as input data. These data were obtained from 52 people and captured under outdoor conditions (in midday and the afternoon) using a webcam. The CNN-VGG was utilized because it is deep and fast enough for both face recognition and facial expression recognition purposes. The results showed that the VGG-f model architecture could overcome the underfitting and overfitting problems stemming from simpler CNN architectures. The testing results showed that the VGG-f model could recognize faces and facial expressions well. The average accuracies achieved in recognizing 104 faces during the day and in the afternoon were 86.5% and 90.4%, respectively. Additionally, the average accuracies achieved in recognizing the four different facial expressions of 52 people were 72% and 74% during the day and at noon, respectively. Recognition errors may have been caused by similarities between images.

In the last decade, long-span continuous beam bridges have been built to meet the need of the totalbridge length requirements for obstacle avoidance. In almost 100 years of development of bridge construction coupled with the development of concrete technology, the maximum bridge length, which was originally about 20 m, currently has reached a length of 50 m or longer with the use of prestressed construction. The longer girder length will reduce the number of pillars, minimize the possibility of damage to the pillars due to local scouring or expand the coverage of the free area under the bridge. This study aimed to evaluate the economical aspect of the use of prestressed girders with length variations of 40 m, 50 m, and 60 m. The analytical calculations were carried out based on the applicable standards of bridge design. The results of the study indicated that the most economical configuration for a long span bridge (200m-400m) is the use of a prestressed girder length of 50 m with a girder cross-section height of 2.10 m. However for 480m bridge lengths, 60m PCI girder lengths tend to start out lower. It can be concluded that for a longer bridge span, a longer PCI girder is also required.

Batik Surakarta has characteristics that not everyone understands, the difference in characteristics lies in the design characteristics, primary colors, and motifs. Surakarta's traditional batik designs and motifs have cultural values that are worth preserving. The combination of the otsu method and the canny method is used to build an application system to improve the results of identifying traditional Surakarta batik images, characteristic features, and shapes of batik patterns. Segmentation is done by partitioning the image into the object area using otsu and determining the boundary of the object using canny to get the characteristics of the image. The results of segmentation processed by selecting and extracting features by selecting quantitative information from existing features using the GLCM method. The characteristics taken are energy, homogeneity, correlation, and contrast. The results of the four characteristics classified into seven Surakarta batik classes, namely Kawung motif, Sido Mukti motif, Truntum motif, Sawat motif, Satrio Manah motif, Parang motif, and Semen Rante motif. The data used are 100 image data, which divided into 70 training data and 30 test data. The results showed that the accuracy rate was 93%.

This paper presents research on synchronizing the data of the Electrooculography (EOG) sensor with eye image data when someone blinks. In this study, we carry out the validation of the vertical blink data synchronization from the EOGV sensor to the blinking eye image. The research uses two modalities to get data validity, whether someone blinks or not. There is ambiguity in the two modalities that have different data rates, and both might also have different false data that should not be recognized as signals or blinking eyelid data the same. Data modality matching and data interpolation are implemented to get complete data following the data duration and to get the pair of both modalities. We use linear regression to obtain a formulation of the synchronous relationship between the EOG blinking data sequence and the number of blinking eyelid image frame data.

Indonesia is known as an archipelago country consisting of various tribes with their cultures and languages. Ancient manuscripts from these various tribes are widely found and need to be preserve. There has been much research on the Javanese script in the Manuscript, such as Javanese script segmentation, Javanese script introduction, and improvement of Javanese characters whose writing has been damage due to the storage process and the length of storage time. This research is focused only on localizing Javanese characters into characters that can be used in recognition of Javanese characters in future research. This study conducted a process of character segmentation of Javanese characters in the ancient Manuscript by using the connected component labeling method, which is applied as an algorithm used to detect interconnected blob regions in a binary image. The accuracy of the results of the segmentation process is quite good at 80.9%.

Orchestra transcription is a complex task in AMT (Automatic Music Transcription) because it contains a mixture of many notations from various instrument sources. In this paper we propose a reliable method for the transcription of notation of the balungan group gamelan orchestra consisting of demung, saron and peking instruments. Orchestra recordings did not initially have a ground truth label. Therefore, the method we propose is divided into two parts. First, constructs ground truth by recognizing each instrument in the orchestra with bandpass filter. Then, detect the pitch notation and its duration using adaptive threshold and frequency threshold. Second, orchestra signals in the time-frequency domain is trained together with the ground truth that has been built using a neural network. Our proposed method has been evaluated and produce significant testing performance with an average accuracy is 0.9658.

This research aims to determine the effect of wimba learning strategies with project based learning models toward student learning outcomes on the excretory system material. This research was conducted in September 2018 until September 2019. The research method used a true experiment with a population of all class XI in one of the State High School of Tasikmalaya City with 8 classes with a total of 288 students. The samples used were 2 classes taken by cluster random sampling, namely class XI mathematics and science 1 as a controller class and class XI mathematics and science 4 as an experimental class. Based on the results of the study, data analysis using the Ancova test and hypothesis testing can be concluded that there is an influence of the wimba learning strategy with project based learning model toward student learning outcomes on the excretory system material in class XI in one of the State High School of Tasikmalaya City 2018/2019.

Unmaned Ground Vehicle (UGV) merupakan teknologi kendaraan darat tanpa awak yang berguna untuk mempermudah pekerjaan manusia dalam berbagai bidang seperti transportasi, aktivitas logistik industri, search and resque, pertahanan dan keamanan, juga beberapa bidang lainnya. Pengendalian attitude menjadi permasalahan karena membutuhkan ketelitian akibat adanya pengaruh kecepatan. Selain itu, bagaimana UGV tersebut mengikuti jalur yang ditentukan juga memerlukan pengendalian attitude yang optimal. Penelitian ini bertujuan untuk merancang dan menguji performa serta untuk mengatahui tingkat keberhasilan dan keakuratan pengendalian UGV menggunakan algoritma Backpropagaion Neural Network. Dari hasil pengujian didapatkan bahwa algoritma ini berhasil mengikuti data uji yang diberikan dengan nilai MSE yang kecil.

This study aims to know the changes that happened in the learning environment caused by the information technology development applied in online learning during the pandemic COVID-19. The rapid changes in information technology have affected many learning strategies and environments, and also improved teacher and student capabilities in information technology. To obtain the solution, this study uses a qualitative descriptive approach as a method. The objects in this research are teachers and students in Aviation Polytechnic of Surabaya, a vocational college that implemented a ruled-boarding school system. The pandemic condition has changed the learning system from face-to-face traditional learning into online learning from their home. The result shows that information technology has successfully shifted the student tendency to do face-to-face learning system. This situation also encourages teachers to be able to choose and use information technology as a learning medium for student learning success. In this study, it was stated that teachers, students, and the Aviation Polytechnic of Surabaya as a vocational school institution need to make some adjustments in learning techniques to accommodate a pandemic learning environment. The learning design allows many models, methods, and media in the online learning system to facilitate the distance learning process to reach certain goals. The implication of this COVID-19 online study-research is a shift in the learning environment caused by the knowledge level of information technology.

This article provides a discussion of the testing and measurement of UHF RFID with distance and facing angle parameters on static and moving state conditions. This study is necessary for implementing RFID technology in the development of autonomous vehicle navigation systems. Navigation systems in autonomous vehicles generally never leave the global positioning system (GPS) as a navigation sensor. The use of GPS independently has weaknesses related to the accuracy, so a navigation system using GPS requires correction of the navigation route based on coordinates, this correction can be done by adding another sensor. The integration of GPS and RFID technology has several advantages besides being cost-effective. Studies that have been carried out enable an autonomous vehicle navigation system to be run by combining data between RFID Reader readings in retrieving location data points marked with RFID tags and coordinate vehicle position data on maps by the GPS which generates route and location information passed by vehicles using the GPS/RFID method localization. Tests and measurements are performed by reading on three types of RFID tags with varying distances and angles of view. The results showed that the best reading distance for RFID tags is at a distance of 4 meters with a reading angle of the RFID Reader at 90 degrees on the z-axis and y-axis. While the best RFID tag performance is the tag on the Passive UHF RFID metal, both for testing in static or moving state condition.

Autonomous vehicles are vehicles that run automatically without a driver. Therefore, the vehicle requires sensors to detect surrounding objects to avoid collisions with other objects or vehicles. A variety of sensors can be used to support the system. One of the sensors used is an ultrasonic sensor that has the reliability and robustness to various light conditions and radio waves, especially when compared to camera sensors, radar and lidar. The recent implementation of ultrasonic sensors in vehicles is limited as a parking guide, so it needs to be developed for further functions, considering that ultrasonic sonar technology has advanced with even greater detection and long distances range. Hence, as a continuation of previous research in ultrasonic sensor characteristics, this paper carries out the application of artificial neural network algorithms that get input in the form of signals that refer to the output signal from the ultrasonic sensors, which have already assembled into a multi-sensor, which is 8 (eight) ultrasonic sensors positioned around the vehicle, two sensors in the front, two sensors in the rear and four sensors in the right and left side of vehicle. The sensors and algorithms will support the autonomous vehicle system, where if the sensors detect the obstructive objects, the system will provide an output in the form of a decision to make the braking order, soft braking, turning left, turning right, or staying run straight when the front sensors do not detect a barrier object. This is done in anticipation of an accident and avoid a collision. Each condition and decision will be determined by which sensor detects the barrier object. Input and output will be simulated using the tool of artificial neural network algorithms, so as to get the most optimal weight and low error rate.

Music is one of the arts of a beauty that cannot be separated from humans. But not all humans can create beautiful music. Therefore, this paper's objective was to design a creative aid to musical composition with minimum human intervention. This paper purpose builds generative music using Multilayer Perceptron (MLP) Neural Networks as architecture to create predicted chord as musical accompaniment. In contrast, the previous work in music generation has mainly been focused on creating subsequent next music patterns. Using the symbolic Indonesian music dataset (note and chord) to train the models. Finally, we predict sequential chords using symbolic notes from the data test. Later, the generated Chords by Neural Network is compared with the original music Indonesian.

Perkembangan mobil listrik sangat pesat seiring dengan semakin berkurangnya sumber energi fosil. Untuk mampu bergerak otomatis identifikasi dan deteksi jalan sangat diperlukan. Namun proses ini sulit dikarenakan jalan yang ada tidak memiliki garis sebagai acuan. Banyak metode yang telah digunakan salah satunya dengan menggunakan Fully Convolutional Networks (FCNs). Metode ini berhasil dalam melakukan identifikasi terhadap jalan yang ada pada kampus Universitas Sriwijaya. Berdasarkan hasil pengujian didapatkan nilai Intersection over Union (IoU) 90%. Sehingga, model yang dihasilkan oleh FCNs dapat digunakan untuk identifikasi jalan yang dilalui. Selain itu parameter lain yang diperhitungkan yaitu nilai akurasi 98,12% pada data latih dan 97,87% pada data uji. Sedangkan error yang dihasilkan sebesar 6 % pada data latih dan 7% pada data uji. Kata kunci: Fully Convolutional Networks (FCNs), Intersection over Union, Jalan kampus, Mobil Listrik ABSTRACT The development of electric cars is very rapid along with the decreasing source of fossil energy. To move automatically, the electric car is needs identification and detection of roads. But this process is difficult because the existing road does not have a line as a reference. Many methods have been used, one of them is using Fully Convolutional Networks (FCNs). This method is successful in identifying existing roads on the Sriwijaya University campus. Based on the results of testing, it obtained Intersection over Union (IoU) value of 90%. So, the model produced by FCNs can be used to identify the path traveled. In addition, other parameters taken into account are the accuracy value of 98.12% in the training data and 97.87% in the test data. While the resulting error of 6% in training data and 7% in test data.

In a text matching similarity task, a model takes two sequence of text as an input and predicts a category or scale value to show their relationship. A developed model is to measure the similarity - one of relationship between those two text. The model is SIAMESE network that implement two copies of same network of CNN, it takes text_1 and text_2 as the inputs respectively for two CNN networks. The output of each CNN network is features vector of the corresponding text input, both outputs are then fed by a loss function to calculate the value of loss (i.e. similarity). This research implemented two types of loss functions, i.e. Triplet loss and Contrastive loss. The usage purpose of these two types of loss functions was to see the influence toward the measurement results of similarity between two text being compared. The metrices used for this comparison are precision, recall, and F1-score. Based on the experimental results done on 1500 pairs of sentences, and varied on the epoch value starting from 10 until 200 with an increment of 10, showed the best result was for epoch value of 180 with precision 0.8004, recall 0.6780, and F1-score 0.6713 for Triplet loss function; and epoch value of 160 with precision 0.6463, recall 0.6440, and F1-score 0.6451 for Contrastive loss function gave the best performance. So that, the Triplet loss function gave better influence than Contrastive loss function in measuring similarity between two given sentences.