RFRuvita Faurina

Ruvita Faurina

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Ruvita Faurina has a total of 23 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Comparative Study of Pre-Trained CNN for Pinkeye Disease Detection in Cattle and Goat Eyes

    by Ruvita Faurina, Muhammad Farchan Al Rahman, Indra Agustian

    Keratoconjunctivitis, commonly known as "pinkeye" disease, significantly impacts the health and productivity of cattle and goats. This study conducts a comparative analysis of three pre-trained convolutional neural network (CNN) models - MobileNetV2, VGG16, and DenseNet - for classifying pinkeye disease images in livestock. The research comprises four phases: data collection, data preparation, modeling, and evaluation. Data collection involves a dataset of 1425 images gathered from various sources. Preprocessing includes resizing, augmentation, and dataset stratification. The study creates two CNN models tailored for cattle and goats using pre-trained CNN. MobileNetV2 consistently demonstrates superior generalization, surpassing 95% accuracy. In contrast, VGG16 and DenseNet201, while achieving higher overall accuracy, show overfitting. MobileNetV2 is identified as the most proficient model for pinkeye disease classification, advancing automated disease diagnosis in cattle and goat farming. Further validation in diverse operational contexts is recommended.

  • Open Access

    Comparison of Convolutional Neural Networks Transfer Learning Models for Disease Classification of Food Crop

    by Ruvita Faurina, Silvia Rahma, Arie Vatresia, Agus Susanto

    Indonesia is an agricultural country with 29% of the workforce working in the agricultural sector, however, farmers' knowledge and practices depend on informal local wisdom based on inherited past practices. Moreover, identifying diseases in plants is difficult to do with human vision so that intelligent technology is needed. In this paper, an architecture of CNN models such as MobileNetV2, ResNetV50, InceptionV3 and DenseNet121 will be built to detect diseases based on leaf images of several crops obtained from the agroai dataset containing multiple crops namely bean, chili, corn, potato, tomato and tea. The model is used through transfer learning for feature extraction of the trained model with imagenet weights, with 4 fully connected layers. Each model for each crop will be compared to get the best model based on the accuracy of training, evaluation and testing. ResNet50 has the best performance for four type of plants, including bean plants with training accuracy of 99.49%, validation of 99.52%, testing of 98.96%, chili plants with training accuracy of 98.03%, evaluation of 98.75%, testing of 100%, tea plants with training accuracy of 99.62%, evaluation of 99.6%, testing of 99.74% and tomato plants with training accuracy of 99.62%, validation of 99.7%, testing of 99.37%. Moreover, MobileNetV3 has the best performance for 2 types of crops that is corn with training accuracy of 99.22%, validation of 99.69%, testing of 99.55%, and potato with training accuracy of 99.62%, evaluation of 99.60%, testing of 99.74%.


  • Open Access

    FISH FRESHNESS PREDICTION WITH CONVOLUTIONAL NEURAL NETWORK METHOD BASED ON FISH EYE IMAGE ANALYSIS

    by Robby Mahendra, Ruvita Faurina

    The potential for fish resources in Bengkulu waters is abundant, but quality must be maintained for safety and selling value. Changes in the skin, eyes, gills and flesh of fish indicate a decrease in quality due to enzyme, chemical and bacterial activity. The process of sorting fish by fishermen or sellers is still often done manually, which is sometimes inaccurate due to limited vision. With advances in computing technology, classification algorithms are needed that can identify and differentiate between fresh fish and non-fresh fish. This research uses a Convolutional Neural Network with DenseNet201, VGG16, and InceptionV3 architecture. The dataset contains 880 Belato Alepes Djedaba fish eye images, with a ratio of 80:15:5 for train, validation, and test. DenseNet201 has the best performance compared to VGG16 and InceptionV3. Accuracy on DenseNet201 test data 98%, InceptionV3 95%, and VGG16 91%. The classification results of the best model using 8 images with various scenarios show that all images were successfully classified 100% correctly. This research makes a contribution to the field of fishery product processing technology which allows fish quality classification to be carried out quickly and accurately, as well as increasing efficiency in ensuring the quality of fish for consumption.


  • Open Access

    Implementasi Metode Naïve Bayes Pada Penentuan Mutu CPO (Crude Palm Oil)

    by Rusdi Efendi, Ruvita Faurina, Tiya Suci Hamimmah

    Kelapa sawit merupakan sumber minyak nabati yang sangat produktif dan menghasilkan minyak nabati tertinggi dibandingkan tanaman perkebunan lainnya menjadikan latar belakang penelitian ini. Tujuan penelitian ini melakukan klasifikasi mutu CPO (Crude Palm Oil) kedalam 2 kelas klasifikasi menggunakan metode Naïve Bayes. Dengan adanya klasifikasi ini diharapkan dapat mengurangi human error dalam dalam penentuan mutu CPO. Parameter yang digunakan dalam peneltian ini adalah kadar asam lemak bebas, kadar air dan kadar kotoran yang sangat berpengaruh terhadap mutu CPO. Dengan menggunakan teknik split data pada saat vaidasi, data dibagi menjadi data latih dan data uji sebesar 70:30. Berdasarkan hasil pengujian menggunakan confussion matrix, penggunaan metode Naïve Bayes terhadap data yang telah diambil pada objek penelitian diperoleh tingkat akurasi sebesar 97.7 % atau termasuk dalam kategori excellent. Sementara nilai Spesifisitas sebesar 95.8 %, nilai recall sebesar 100% dan Presisi sebesar 95.4%.


  • Open Access

    Deep learning pest detection on Indonesian red chili pepper plant based on fine-tuned YOLOv5

    by Indra Agustian, Ruvita Faurina, Sahrial Ihsani Ishak, Ferzha Putra Utama, Kusmea Dinata Dinata, Novalio Daratha

    .This research developed a pest detection model for Indonesian red chili pepper based on fine-tuned YOLOv5. Indonesian red chili pepper is the third largest vegetable commodity produced in Indonesia. Pest attacks disrupt the quantity and quality of crop yields. To control pests effectively, it is necessary to detect the type of pest correctly. A viable solution is to leverage computer vision and deep learning technologies. However, no previous studies have developed a pest detection model for Indonesian red chili pepper based on this technology. YOLOv5 is a variant of the YOLO object detection algorithm, which has major advantages in terms of computation cost and execution speed. The dataset comprises 4,994 image files collected from a chili plantation in Bengkulu province, Indonesia, covering 4 different classes and a total of 10,683 pests. The image is 1216 x1216 px with the smallest, largest, and average object dimensions of 2%, 35%, and 4% of the image dimensions. The training model used is fine-tuning YOLOv5s with variations of patience as an early stop parameter of 100, 200, and 300. The evaluation of the trained model is based on train loss, validation loss, and mAP@0.5:0.95, the best-trained model is the 445th epoch on patience 100 with the best confidence value of 0.321 and the highest TF1 of 0.74. From the best-trained model testing on the test dataset, the mAP@0.5 performance for all classes is 81.3%. The model not only detected large pests but was also able to detect objects that were small in size compared to the image size. The best-trained model's best mAP@0.5 performance and speed are 82.6% and 20 ms/image, or 50 fps on NVIDIA P100 GPU.


  • Open Access

    Sentiment Analysis Komentar Berbahasa Bengkulu Menggunakan Long Short-Term Memory (LSTM)

    by Safitri Nurhaeni, Ruvita Faurina, Ferzha Putra Utama, Kurnia Anggriani

    Sentiment information is one type of information that can be obtained from social media. Sentiments can be interpreted as opinions and views of society which contain feelings. To analyze the value of sentiment whether the sentiment is a sentiment that tends to be neutral, negative, or positive, sentiment analysis can be used. language has its characteristics and uniqueness, Bengkulu language is no exception, because of this, it is necessary to model sentiment analysis for various languages. Sentiment modeling for the Bengkulu language is not yet available, therefore a sentiment analysis model for the Bengkulu language is developed by applying Long Short-Term Memory (LSTM), and architectural experiments for Long Short-Term Memory (LSTM) are carried out to obtain an architectural sentiment analysis model that produces the best value. The data used in the study amounted to 24,000 Bengkulu-language comments received from social media Instagram, Twitter, and Youtube. Experimental research 1 produces the best accuracy value compared to the results of testing in other experiments, with an accuracy value of 0.87 a precision value of 0.80, a recall value of 0.82, and an F1 score of 0.81 Keywords: Information, sentiment, Long Short- Term Memory (LSTM), Bengkulu Language, architecture, social media.


  • Open Access

    Implementasi Deep Feed-Forward Neural Network pada Perancangan Chatbot Berbasis Web di UPPIK RSUD M. YUNUS

    by Ruvita Faurina, M. Jumli Gazali, Icha Dwi Aprilia Herani

    Nilai precision sebesar 0,99, recall sebesar 0,98, dan f1-score sebesar 0,99.


  • Open Access

    Image captioning to aid blind and visually impaired outdoor navigation

    by Ruvita Faurina, Anisa Jelita, Arie Vatresia, Indra Agustian

    Artificial intelligence technology has dramatically improved the quality of services for human needs, one of which is technology to improve the quality of services for the blind and visually impaired, particularly technology that can help them understand visual sights to facilitate navigation in their daily lives. This study developed an image captioning model to aid the blind and visually impaired in outdoor navigation. The image captioning model employs the encoder-decoder method, with the convolutional neural network (CNN) feature extraction and attention layer as encoders and the long short-term memory (LSTM) as decoders. ResNet101 and ResNet152 are used in the encoder to extract image features. The results of the extraction and caption are forwarded to the attention layer and the LSTM network. The attention layer uses the Bahdanau attention mechanism. The accuracy of the model is calculated using the bilingual evaluation understudy score (BLEU), metric for evaluation of translation with explicit ordering (METEOR) and recall-oriented understudy for gisting evaluation-longest common subsequence (ROUGE-L). ResNet101 performed the best on BLEU-4, scoring 91.811% and 94.0337% in the METEOR evaluation. The captioning results show that the model is quite successful in displaying a simple caption that is suitable for each image.


  • Open Access

    Comparative study of ensemble deep learning models to determine the classification of turtle species

    by Ruvita Faurina, Andang Wijanarko, Aknia Faza Heryuanti, Sahrial Ihsani Ishak, Indra Agustian

    Sea turtles are reptiles listed on the international union for conservation of nature (IUCN) red list of threatened species and the convention on international trade in endangered species of wild fauna and flora (CITES) Appendix I as species threatened with extinction. Sea turtles are nearly extinct due to natural predators and people who are frequently incorrect or even ignorant in determining which turtles should not be caught. The aim of this study was to develop a classification system to help classify sea turtle species. Therefore, the ensemble deep learning of convolutional neural network (CNN) method based on transfer learning is proposed for the classification of turtle species found in coastal communities. In this case, there are five well-known CNN models (VGG-16, ResNet-50, ResNet-152, Inception-V3, and DenseNet201). Among the five different models, the three most successful were selected for the ensemble method. The final result is obtained by combining the predictions of the CNN model with the ensemble method during the test. The evaluation result shows that the VGG16 - DenseNet201 ensemble is the best ensemble model, with accuracy, precision, recall, and F1-Score values of 0.74, 0.75, 0.74, and 0.76, respectively. This result also shows that this ensemble model outperforms the original model.

  • Open Access

    Comparative study of ensemble deep learning models to determine the classification of turtle species

    by Ruvita Faurina, Andang Wijanarko, Aknia Faza Heryuanti, Sahrial Ihsani Ishak, Indra Agustian

    Sea turtles are reptiles listed on the international union for conservation of nature (IUCN) red list of threatened species and the convention on international trade in endangered species of wild fauna and flora (CITES) Appendix I as species threatened with extinction. Sea turtles are nearly extinct due to natural predators and people who are frequently incorrect or even ignorant in determining which turtles should not be caught. The aim of this study was to develop a classification system to help classify sea turtle species. Therefore, the ensemble deep learning of convolutional neural network (CNN) method based on transfer learning is proposed for the classification of turtle species found in coastal communities. In this case, there are five well-known CNN models (VGG-16, ResNet-50, ResNet-152, Inception-V3, and DenseNet201). Among the five different models, the three most successful were selected for the ensemble method. The final result is obtained by combining the predictions of the CNN model with the ensemble method during the test. The evaluation result shows that the VGG16 - DenseNet201 ensemble is the best ensemble model, with accuracy, precision, recall, and F1-Score values of 0.74, 0.75, 0.74, and 0.76, respectively. This result also shows that this ensemble model outperforms the original model.