3 research outputs found
RANCANG BANGUN PURWARUPA MOBIL OTONOM MELALUI TEKNIK VISI KOMPUTER DAN DEEPLEARNING DENGAN KENDALI PURE PURSUIT
No full textSelf-Supervised Bi-Pipeline Learning Approach for High Interpretation of Breast Thermal Images
No full textThe image quality supports a high accuracy rate of medical image diagnosis using computer vision. Digital thermal images resulting from the thermal device usually suffer from many watermarks that may lower the neural network learning performance. Thus, providing only the region of interest (RoI) of the breast area from the breast thermal images for early breast cancer detection is an important task. The goal of our work are to develop a deep learning (DL) model for taking the RoI of the breast thermal images, built a self-supervised DL model to classify the breast thermal images into healthy and cancer categories, and integrated these two models as end-to-end bi-pipeline model for breast thermal image recognition. The segmentation model was built using attention U-Net with residual recurrent network called R2AU-Net, and the classification model was built using self-supervised learning consisting of the Simple Framework for Contrastive Learning of Visual Representations (SimCLR) and ResNet50. These networks were trained using unlabelled limited breast thermal datasets to allow more comprehensive learning. The result shows that proposed self-supervised bi-pipeline model can take the RoI with an accuracy rate of 98.63% and classify the breast thermal images with a top-1 accuracy rate of 84.37% and top-5 accuracy rate of 96.87%. In addition, the bi-pipeline model implementation using a central processing unit shows that the model required only about 4 seconds for segmentation and classification tasks. These findings indicate that the bi-pipeline model can effectively aid the interpretation of unlabeled breast thermal images
Improving Bi-LSTM for High Accuracy Protein Sequence Family Classifier
No full textThe primary nutrient that is crucial for identifying biochemical processes and biological norms in living cells is protein. Proteins are usually centered around one or a few functions which are defined by their family type. Hence, identification and classification are needed to separate the proteins according to their structure and families. In this work, we built a model to classify families of protein sequences. We used the protein sequences dataset consists of various macromolecules of biological significance. The classifier is built up using deep learning of Bi-LSTM. We began the research by collecting the dataset from the Protein Data Bank of the Research Collaboratory for Structural Bioinformatics, pre-processing the data using tokenizing, and modeling the classifier based on deep learning network of Bi-LSTM. As we get the best accuracy rate of the trained model, we figure out the model performance using the evaluation metrics of learning curve, accuracy rate, and loss. The results show that Deep Bi-LSTM provides excellent performance with fit learning curve, 99% accuracy rate, and 0.042 loss
