Segment Anything Model (SAM) is a model capable of performing object segmentation in images without requiring any additional training. Although the segmentation produced by SAM lacks high precision, this model holds interesting potential for more accurate segmentation tasks. In this study, we propose a Post-Processing method called Conditional Matting 4 (CM4) to enhance high-precision object segmentation, including prominent, occluded, and complex boundary objects in the segmentation results from SAM. The proposed CM4 Post-Processing method incorporates the use of morphological operations, DistilBERT, InSPyReNet, Grounding DINO, and ViTMatte. We combine these methods to improve the object segmentation produced by SAM. Evaluation is conducted using metrics such as IoU, SAD, MAD, Grad, and Conn. The results of this study show that the proposed CM4 Post-Processing method successfully improves object segmentation with a SAD evaluation score of 20.42 (a 27% improvement from the previous study) and an MSE evaluation score of 21.64 (a 45% improvement from the previous study) compared to the previous research on the AIM-500 dataset. The significant improvement in evaluation scores demonstrates the enhanced capability of CM4 in achieving high precision and overcoming the limitations of the initial segmentation produced by SAM. The contribution of this research lies in the development of an effective CM4 Post-Processing method for enhancing object segmentation in images with high precision. This method holds potential for various computer vision applications that require accurate and detailed object segmentation.

Technological developments go hand in hand with advances in digital messaging. In protecting the confidentiality of the message, it is necessary to double secure the data. This security can be done with a combination of steganography and cryptographic techniques. Steganography algorithm which is a technique for hiding messages well, one of which is Least Significant Bit (LSB). The LSB algorithm is a simple method because it only converts the value of the last bit in a message with the inserted message bit, which is a convenience of the LSB algorithm, but it becomes vulnerable to message theft attacks if not combined with other algorithms for security. So it is necessary to increase security. This research developed a combination method of LSB algorithm for steganography technique with Hill Cipher algorithm for cryptographic technique, Hill Cipher was developed with shifting (shifting) 2 (two) characters. With the development of this method, hackers will find it difficult to crack messages, and is expected to improve the performance of the algorithm in affecting image quality and travel time in running the algorithm. The results of this study will be tested using several evaluation tools MSE, PSNR, BER, CER, AE, and Entropy. With the development of this method, hackers will find it difficult to decipher messages, and from the results of this experiment has been able to improve the performance of the algorithm in maintaining image quality and can shorten travel time in running the algorithm.

Online lending services such as Peer to Peer (P2P) loans provide convenience for lenders to transact directly without involving banks as intermediaries. Identifying potential loan recipients who are at risk of default is a crucial step in preventing financial losses, as lenders are responsible for default risk. However, predicting default risk becomes a challenge when P2P lending datasets have various complex features. Some features in P2P lending are redundant, while others do not significantly contribute to an effective solution. Therefore, feature selection is an important process to choose a relevant subset of features from input or target data. Traditional feature selection methods often fail to provide optimal results. A better approach is to use heuristic search algorithms capable of finding suboptimal feature subsets. We employ the Grey Wolf Optimization (GWO) technique, inspired by the hierarchy of leadership and grey wolf hunting mechanisms. Combined with Random Forest (RF), which has limitations in classifying data with very high dimensions, our GWO+RF combination has proven to enhance classification performance better than previous research. It achieves an accuracy score of 97.31%, compared to previous research with scores of only 67.72% for RBM+RF, 64% for Binary PSO+ERT, and 92% for GA+RF.

Flowers are plants that had many types and often found around. But because the many types of flowers, sometimes difficult to distinguish the type from one flower to another. Therefore, in this study, will discuse about the process of identification and classification of flower types, namely daisy, dandelion, rose, sunflower and tulip. The data that would used in this research is image data that consisting of 764 daisy images, 1052 dandelion images, 784 rose images, 733 sunflower images and 984 tulip images. From the total images used, would be divided again into 60% training data, 30% testing data and 10% validation data that would been used to train and evaluate the CNN model. In this study, the classification process would using transfer learning CNN method using the DenseNet and NasNetLarge architectures, which later from these two architectures would compare to find which architecture is best for classifying flower types. The results that obtained after testing in this study are in the flower classification process using the DenseNet architecture to get a test accuracy of 89% and using the NasLargeNet architecture to get a test accuracy of 86%.

Health is a very important thing in life. Therefore, to maintain health, we need adequate rest. Without adequate rest, the body will not be healthy and fit. In this study, a person's sleep disorder prediction will be made based on their lifestyle and work. The predictions made will classify sleep disorders that are absent, sleep apnea and insomnia from certain lifestyles and work. The methods used to make predictions are decision tree classifier, random forest classifier and naïve Bayes classifier. The test was carried out using a total of 375 data which was broken down into 70% training data and 30% testing data. The results obtained after testing with test data are by using the decision tree classifier algorithm to get an accuracy of 89.431%, using the random forest classifier algorithm to get an accuracy of 90.244% and by using the naïve Bayes classifier algorithm to get an accuracy of 86.992%.

In this comprehensive study, a meticulous analysis of the application of Convolutional Neural Network (CNN) methodologies in the classification of Covid-19 and non-Covid-19 cases was conducted. Leveraging diverse optimization techniques such as RMS, SGD, and Adam, the research systematically evaluated the performance of the CNN model in accurately discerning intricate patterns and distinct features associated with Covid-19 pathology. the implementation of the RMS and Adam optimization methods resulted in the highest accuracy levels, with both models achieving an impressive 98% accuracy in the classification of Covid-19 and non-Covid-19 cases. Leveraging the robust capabilities of these optimization techniques, the study successfully demonstrated the effectiveness of the RMS and Adam models in enhancing the precision and reliability of the Convolutional Neural Network (CNN) for the accurate identification and differentiation of Covid-19 patterns within the medical imaging datasets. The notable achievement of 98% accuracy further emphasizes the potential of these optimization methods in advancing the capabilities of CNN-based diagnostic tools, thus contributing significantly to the ongoing efforts in Covid-19 diagnosis and management.  

This research focuses on car classification and the use of the ResNet-50 neural network architecture to improve the accuracy and reliability of car detection systems. Indonesia, as one of the countries with high daily mobility, has a majority of the population using cars as the main mode of transportation. Along with the increasing use of cars in Indonesia, many automotive industries have built factories in this country, so the cars used are either local or imported. The importance of car classification in traffic management is a major concern, and vehicle make and model recognition plays an important role in traffic monitoring. This study uses the Vehicle images dataset which contains high-resolution images of cars taken from the highway with varying viewing angles and frame rates. This data is used to analyze the best- selling car brands and build car classifications based on output or categories that consumers are interested in. Digital image processing methods, machine learning, and artificial neural networks are used in the development of automatic and real-time car detection systems.The ResNet-50 architecture was chosen because of its ability to overcome performance degradation problems and study complex and abstract features from car images. Residual blocks in the ResNet architecture allow a direct flow of information from the input layer to the output layer, overcoming the performance degradation problem common in neural networks. In this paper, we explain the basic concepts of ResNet-50 in car detection and popular techniques such as optimization, augmentation, and learning rate to improve performance and accuracy. in this study, it is proved that ResNet has a fairly high accuracy of 95%, 92% precision, 93% recall, and 92% F1-Score.

The use of IoT sensors in technology is a fascinating research subject due to the possibility of real-time monitoring of water usage and temperature effects. Developing tools and systems that make observations of water usage through Max6675 sensors can optimize data collection and processing through remote monitoring. Utilizing the IoT technique, the ESP8266 WeMOS D1 R2 microcontroller and the Max6675 temperature sensor are utilized to regulate the water pump via a relay. This approach enhances water sustainability for plant assumptions based on the predetermined temperature. The device activates the water pump when the temperature surpasses the standard level at the observation site. The outcome indicated that this device operates accordingly with a 100% success rate. The device was determined to function effectively by activating the water pump based on temperature. Additionally, it can provide real-time monitoring data and process sensor data for analysis.

Film sentiment analysis is a process for evaluating a sentiment value that exists in film reviews, so that positive or negative responses from films can be identified. In this study, a sentiment analysis will be carried out on film reviews on IMBD. The analysis was carried out to find out which reviews were positive and negative from film critics. The method used to carry out sentiment analysis in this study is review analysis and processing with TF-IDF and a positive or negative prediction process based on reviews that have been processed using a logistic regression algorithm and support vector classification. The data to be used is film reviews on IMBD, which consists of 2000 data, which is divided into 1000 positive data and 1000 negative data. Which is where the data will be preprocessed first and split with a percentage of 70% training data and 30% testing data. In the prediction process using the logistic regression algorithm, obtaining a test accuracy of 80.61%. While the prediction process using the support vector classification algorithm obtains a test accuracy of 82.42%.

Watermarking is a branch of the data hiding technique. Watermarking is a technique used to insert a copyright label on an image, so that the copyright of the image can be protected. Discrete Cosine Transform (DCT) and Discrete Wavelet Transform (DWT) are techniques that can be used to watermark. In this study, the Discrete Cosine Transform and Discrete Wavelet Transform methods will be used to watermark images to 5 different host images. In the tests carried out, watermarking techniques will be compared using DCT, DWT, DCT-DWT combination and DWT-DCT combination. The results obtained in this study were the highest PSNR value obtained at 41.931, the highest SSIM obtained 0.99515, the highest entropy was also obtained at 7.4186, The best UACI value is 0.0071158 and the best NCPR value is obtained at 93.9068% then, for the best CC value is obtained at 0.99953. As well as the NCC value, the value obtained is the same all in each test, namely with a value of 1.