In this paper, we aim specifically to address suspicious activities of ATM CCTV footage such as when a consumer left his wallet, consumer shows confused gesture because the ATM dispenses no transaction receipt, ATM card is swallowed, etc. We proposed 3D CNN for activity recognition. We also proposed a tailored algorithm to detect the time frame in which the activity happens in the untrimmed input video.[ This paper presents an approach to temporal human activity detection using the proposal then classification framework, which is one of the frameworks for temporal activity detection. The goal of this research is to detect and recognize certain activities at the ATM. We propose an activity detection method using a 3D convolutional neural network (3D CNN). Our proposed method achieved performance with the accuracy score of 93.94%, a precision of 96.36%, a recall of 93.94%, and an f-score of 93.69%.]

This work applies a hierarchical transfer learning to implement deep neural network (DNN)-based multilingual text-to-speech (TTS) for low-resource languages. DNN-based system typically requires a large amount of training data. In recent years, while DNN-based TTS has made remarkable results for high-resource languages, it still suffers from a data scarcity problem for low-resource languages. In this article, we propose a multi-stage transfer learning strategy to train our TTS model for low-resource languages. We make use of a high-resource language and a joint multilingual dataset of low-resource languages. A pre-trained monolingual TTS on the high-resource language is fine-tuned on the low-resource language using the same model architecture. Then, we apply partial network-based transfer learning from the pre-trained monolingual TTS to a multilingual TTS and finally from the pre-trained multilingual TTS to a multilingual with style transfer TTS. Our experiment on Indonesian, Javanese, and Sundanese languages show adequate quality of synthesized speech. The evaluation of our multilingual TTS reaches a mean opinion score (MOS) of 4.35 for Indonesian (ground truth = 4.36). Whereas for Javanese and Sundanese it reaches a MOS of 4.20 (ground truth = 4.38) and 4.28 (ground truth = 4.20), respectively. For parallel style transfer evaluation, our TTS model reaches an F0 frame error (FFE) of 9.08%, 10.13%, and 8.43% for Indonesian, Javanese, and Sundanese, respectively. The results indicate that the proposed strategy can be effectively applied to the low-resource languages target domain. With a small amount of training data, our models are able to learn step by step from a smaller TTS network to larger networks, produce intelligible speech approaching the real human voice, and successfully transfer speaking style from a reference audio.

Leptomeningeal metastasis is an indication of the malignancy that occurs in leukemia patients. Although it only has a 5-10% portion caused the leukemia patient to relapse, the abnormality is the basis in determining the best treatment given to them. Leptomeningeal metastasis are better detected by using Magnetic Resonance Imaging (MRI) because of their high sensitivity in neuraxis images. High ability to see and analyze is needed for a radiologist in reading the Brain MRI results of leukemia patients with suspect leptomeningeal metastasis. Therefore, the classification will take a long time and allow for the misreading of the results. In this experiment, we used a dataset from the Brain MRI of leukemia patients of Dharmais Cancer Hospital. We implemented the proposed method in performing the leptomeningeal metastasis segmentation. The preprocessing image applied for sharpening and removing unwanted noises in the image using the Median Filter. A hybrid semi-automated skull stripping was also developed to improve the accuracy of the segmentation. Then Fuzzy C-Means is used to segment the abnormalities and reach an average evaluation performance at 49.1% Jaccard Index.

This research introduces least square adversarial autoencoder (LSAA)–an autoencoder that is able to reconstruct data and also generate data that has characteristics similar to data distribution from the prior distribution. LSAA uses least square generative adversarial network loss function on its discriminator. LSAA minimizes Pearson χ 2 divergence between the latent variable distribution and the prior distribution. In this research, a Python program is developed to model LSAA by utilizing MNIST data set and FashionMNIST data set. The program is implemented using PyTorch. All of the programming activities are carried out in the cloud environment provided by the Tokopedia-Universitas Indonesia AI Center, using DGX-1 (GPU Tesla V100) as its computing resource. The experimental results show that the mean squared error of LSAA for MNIST data set and FashionMNIST data set are 0.0080 and 0.0099, respectively. Furthermore, the Frechet Inception Distance score ́ of LSAA for MNIST data set and FashionMNIST data set are 11.1280 and 27.5737, respectively. These results indicate that the least square adversarial autoencoder is able to reconstruct the image properly and also able to generate images similar to the training samples.

Completeness is a crucial data quality aspect that deals with the question: do we have all the data we need? The lack of awareness on the completeness state of a knowledge graph (KG) may result in bias or even falsity for any decisions made based on the KG. Given a KG, one may be wondering how its completeness may vary across different topics. In this paper, we present ProWD, a framework and tool for profiling the completeness of Wikidata, a central KG on the (Semantic) Web that is open and free to use. ProWD measures the degree of completeness based on the Class-Facet-Attribute (CFA) profiles. A class denotes a collection of entities, which can be of multiple facets, allowing attribute completeness to be analyzed and compared, e.g., how does the completeness of the attribute "educated at" and "date of birth" compare between male, German computer scientists, and female, Indonesian computer scientists? ProWD generates summaries and visualizations for such analysis, giving insights into the KG completeness. ProWD is available online at http://prowd.id.

Ontology alignment has been studied for over a decade, and over that time many alignment systems and methods have been developed by researchers in order to find simple 1-to-1 equivalence matches between two ontologies. However, very few alignment systems focus on finding complex correspondences. One reason for this limitation may be that there are no widely accepted alignment benchmarks that contain such complex relationships. In this paper, we propose a real-world dataset from the GeoLink project as a potential complex alignment benchmark. The dataset consists of two ontologies, the GeoLink Base Ontology (GBO) and the GeoLink Modular Ontology (GMO), as well as a manually created reference alignment, that was developed in consultation with domain experts from different institutions. The alignment includes 1:1, 1:n, and m:n equivalence and subsumption correspondences, and is available in both Expressive and Declarative Ontology Alignment Language (EDOAL) and rule syntax.

In our experience, some ontology users find it much easier to convey logical statements using rules rather than OWL (or description logic) axioms. Based on recent theoretical developments on transformations between rules and description logics, we develop ROWL, a Protege plugin that allows users to enter OWL axioms by way of rules; the plugin then automatically converts these rules into OWL DL axioms if possible, and prompts the user in case such a conversion is not possible without weakening the semantics of the rule.