AI, Volume 2, Issue 2 (June 2021) – 9 articles

Artificial intelligence (AI) has brought lots of excitement to our day-to-day lives. Some examples are spam email detection, language translation, etc. Baby monitoring devices are being used to send video data of the baby to the caregiver’s smartphone. However, the automatic understanding of the data was not implemented in most of these devices. In this research, AI and image processing techniques were developed to automatically recognize unwanted situations that the baby was in. The monitoring device automatically detected: (a) whether the baby’s face was covered due to sleeping on the stomach; (b) whether the baby threw off the blanket from the body; (c) whether the baby was moving frequently; (d) whether the baby’s eyes were opened due to awakening. The device sent notifications and generated alerts to the caregiver’s smartphone whenever one or more of these situations occurred. Thus, the caregivers were not required to monitor the baby at regular intervals. They were notified when their attention was required. The device was developed using NVIDIA’s Jetson Nano microcontroller. A night vision camera and Wi-Fi connectivity were interfaced. Deep learning models for pose detection, face and landmark detection were implemented in the microcontroller. A prototype of the monitoring device and the smartphone app were developed and tested successfully for different scenarios. Compared with general baby monitors, the proposed device gives more peace of mind to the caregivers by automatically detecting un-wanted situations. Full article

This paper analyses the contribution of residual network (ResNet) based convolutional neural network (CNN) architecture employed in two tasks related to plant phenotyping. Among the contemporary works for species recognition (SR) and infection detection of plants, the majority of them have performed experiments on balanced datasets and used accuracy as the evaluation parameter. However, this work used an imbalanced dataset having an unequal number of images, applied data augmentation to increase accuracy, organised data as multiple test cases and classes, and, most importantly, employed multiclass classifier evaluation parameters useful for asymmetric class distribution. Additionally, the work addresses typical issues faced such as selecting the size of the dataset, depth of classifiers, training time needed, and analysing the classifier’s performance if various test cases are deployed. In this work, ResNet 20 (V2) architecture has performed significantly well in the tasks of Species Recognition (SR) and Identification of Healthy and Infected Leaves (IHIL) with a Precision of 91.84% and 84.00%, Recall of 91.67% and 83.14% and F1 Score of 91.49% and 83.19%, respectively. Full article

COVID-19 has been a great challenge for humanity since the year 2020. The whole world has made a huge effort to find an effective vaccine in order to save those not yet infected. The alternative solution is early diagnosis, carried out through real-time polymerase chain reaction (RT-PCR) tests or thorax Computer Tomography (CT) scan images. Deep learning algorithms, specifically convolutional neural networks, represent a methodology for image analysis. They optimize the classification design task, which is essential for an automatic approach with different types of images, including medical. In this paper, we adopt a pretrained deep convolutional neural network architecture in order to diagnose COVID-19 disease from CT images. Our idea is inspired by what the whole of humanity is achieving, as the set of multiple contributions is better than any single one for the fight against the pandemic. First, we adapt, and subsequently retrain for our assumption, some neural architectures that have been adopted in other application domains. Secondly, we combine the knowledge extracted from images by the neural architectures in an ensemble classification context. Our experimental phase is performed on a CT image dataset, and the results obtained show the effectiveness of the proposed approach with respect to the state-of-the-art competitors. Full article

Current food crisis predictions are developed by the Famine Early Warning System Network, but they fail to classify the majority of food crisis outbreaks with model metrics of recall (0.23), precision (0.42), and f1 (0.30). In this work, using a World Bank dataset, classical and neural network (NN) machine learning algorithms were developed to predict food crises in 21 countries. The best classical logistic regression algorithm achieved a high level of significance (p < 0.001) and precision (0.75) but was deficient in recall (0.20) and f1 (0.32). Of particular interest, the classical algorithm indicated that the vegetation index and the food price index were both positively correlated with food crises. A novel method for performing an iterative multidimensional hyperparameter search is presented, which resulted in significantly improved performance when applied to this dataset. Four iterations were conducted, which resulted in excellent 0.96 for metrics of precision, recall, and f1. Due to this strong performance, the food crisis year was removed from the dataset to prevent immediate extrapolation when used on future data, and the modeling process was repeated. The best “no year” model metrics remained strong, achieving ≥0.92 for recall, precision, and f1 while meeting a 10% f1 overfitting threshold on the test (0.84) and holdout (0.83) datasets. The year-agnostic neural network model represents a novel approach to classify food crises and outperforms current food crisis prediction efforts. Full article

Matching surfaces is a challenging 3D Computer Vision problem typically addressed by local features. Although a plethora of 3D feature detectors and descriptors have been proposed in literature, it is quite difficult to identify the most effective detector-descriptor pair in a certain application. Yet, it has been shown in recent works that machine learning algorithms can be used to learn an effective 3D detector for any given 3D descriptor. In this paper, we present a performance evaluation of the detector-descriptor pairs obtained by learning a 3D detector for the most popular 3D descriptors. Purposely, we address experimental settings dealing with object recognition and surface registration. Our results show how pairing a learned detector to a learned descriptors like CGF leads to effective local features when pursuing object recognition (e.g., 0.45 recall at 0.8 precision on the UWA dataset), while there is not a clear performance gap between CGF and effective hand-crafted features like SHOT for surface registration (0.18 average precision for the former versus 0.16 for the latter). Full article

In this study we present an autonomous grasping system that uses a vision-guided hand–eye coordination policy with closed-loop vision-based control to ensure a sufficient task success rate while maintaining acceptable manipulation precision. When facing a diversity of tasks with complex environments, an autonomous robot should use the concept of task precision, including the accuracy of perception and precision of manipulation, as opposed to just the grasping success rate typically used in previous works. Task precision combines the advantages of grasping behaviors observed in humans and a grasping method applied in existing works. A visual servoing approach and a subtask decomposition strategy are proposed here to obtain the desired level of task precision. Our system performs satisfactorily on a tangram puzzle task. The experiments highlight the accuracy of perception, precision of manipulation, and robustness of the system. Moreover, the system is of great significance for improving the adaptability and flexibility of autonomous robots. Full article

We propose a method that can perform one-class classification given only a small number of examples from the target class and none from the others. We formulate the learning of meaningful features for one-class classification as a meta-learning problem in which the meta-training stage repeatedly simulates one-class classification, using the classification loss of the chosen algorithm to learn a feature representation. To learn these representations, we require only multiclass data from similar tasks. We show how the Support Vector Data Description method can be used with our method, and also propose a simpler variant based on Prototypical Networks that obtains comparable performance, indicating that learning feature representations directly from data may be more important than which one-class algorithm we choose. We validate our approach by adapting few-shot classification datasets to the few-shot one-class classification scenario, obtaining similar results to the state-of-the-art of traditional one-class classification, and that improves upon that of one-class classification baselines employed in the few-shot setting. Full article

As the amount of scientific information increases steadily, it is crucial to improve fast-reading comprehension. To grasp many scientific articles in a short period, artificial intelligence becomes essential. This paper aims to apply artificial intelligence methodologies to examine broad topics such as enterprise architecture in scientific articles. Analyzing abstracts with latent dirichlet allocation or inverse document frequency appears to be more beneficial than exploring full texts. Furthermore, we demonstrate that t-distributed stochastic neighbor embedding is well suited to explore the degree of connectivity to neighboring topics, such as complexity theory. Artificial intelligence produces results that are similar to those obtained by manual reading. Our full-text study confirms enterprise architecture trends such as sustainability and modeling languages. Full article

In line with the growing trend of using machine learning to help solve combinatorial optimisation problems, one promising idea is to improve node selection within a mixed integer programming (MIP) branch-and-bound tree by using a learned policy. Previous work using imitation learning indicates the feasibility of acquiring a node selection policy, by learning an adaptive node searching order. In contrast, our imitation learning policy is focused solely on learning which of a node’s children to select. We present an offline method to learn such a policy in two settings: one that comprises a heuristic by committing to pruning of nodes; one that is exact and backtracks from a leaf to guarantee finding the optimal integer solution. The former setting corresponds to a child selector during plunging, while the latter is akin to a diving heuristic. We apply the policy within the popular open-source solver SCIP, in both heuristic and exact settings. Empirical results on five MIP datasets indicate that our node selection policy leads to solutions significantly more quickly than the state-of-the-art precedent in the literature. While we do not beat the highly-optimised SCIP state-of-practice baseline node selector in terms of solving time on exact solutions, our heuristic policies have a consistently better optimality gap than all baselines, if the accuracy of the predictive model is sufficient. Further, the results also indicate that, when a time limit is applied, our heuristic method finds better solutions than all baselines in the majority of problems tested. We explain the results by showing that the learned policies have imitated the SCIP baseline, but without the latter’s early plunge abort. Our recommendation is that, despite the clear improvements over the literature, this kind of MIP child selector is better seen in a broader approach to using learning in MIP branch-and-bound tree decisions. Full article