Current Issue : October - December Volume : 2018 Issue Number : 4 Articles : 5 Articles
In the current discussions about ââ?¬Å?artificial intelligenceââ?¬Â (AI) and ââ?¬Å?singularityââ?¬Â, both labels\nare used with several very different senses, and the confusion among these senses is the root of many\ndisagreements. Similarly, although ââ?¬Å?artificial general intelligenceââ?¬Â (AGI) has become a widely used\nterm in the related discussions, many people are not really familiar with this research, including\nits aim and status. We analyze these notions, and introduce the results of our own AGI research.\nOur main conclusions are that: (1) it is possible to build a computer system that follows the same\nlaws of thought and shows similar properties as the human mind, but, since such an AGI will have\nneither a human body nor human experience, it will not behave exactly like a human, nor will it be\nââ?¬Å?smarter than a humanââ?¬Â on all tasks; and (2) since the development of an AGI requires a reasonably\ngood understanding of the general mechanism of intelligence, the systemââ?¬â?¢s behaviors will still be\nunderstandable and predictable in principle. Therefore, the success of AGI will not necessarily lead\nto a singularity beyond which the future becomes completely incomprehensible and uncontrollable....
Traditional image-centered methods of plant identification could be confused due to various views, uneven illuminations, and\ngrowth cycles. To tolerate the significant intraclass variances, the convolutional recurrent neural networks (C-RNNs) are proposed\nfor observation-centered plant identification to mimic human behaviors. The C-RNN model is composed of two components: the\nconvolutional neural network (CNN) backbone is used as a feature extractor for images, and the recurrent neural network (RNN)\nunits are built to synthesizemultiviewfeatures fromeach image for final prediction. Extensive experiments are conducted to explore\nthe best combination of CNN and RNN. All models are trained end-to-end with 1 to 3 plant images of the same observation by\ntruncated back propagation through time. The experiments demonstrate that the combination of MobileNet and Gated Recurrent\nUnit (GRU) is the best trade-off of classification accuracy and computational overhead on the Flavia dataset. On the holdout test\nset, the mean 10-fold accuracy with 1, 2, and 3 input leaves reached 99.53%, 100.00%, and 100.00%, respectively. On the BJFU100\ndataset, the C-RNN model achieves the classification rate of 99.65% by two-stage end-to-end training. The observation-centered\nmethod based on the C-RNNs shows potential to further improve plant identification accuracy....
Despite the diversity of electric wheelchairs, many people with physical limitations and seniors have difficulty using their standard\njoystick. As a result, they cannot meet their needs or ensure safe travel. Recent assistive technologies can help to give them\nautonomy and independence. This work deals with the real-time implementation of an artificial intelligence device to overcome\nthese problems. Following a review of the literature from previous work, we present the methodology and process for\nimplementing our intelligent control system on an electric wheelchair. The system is based on a neural algorithm that\novercomes problems with standard joystick maneuvers such as the inability to move correctly in one direction. However, this\nimplies the need for an appropriate methodology to map the position of the joystick handle. Experiments on a real wheelchair\nare carried out with real patients of the Mohamed Kassab National Institute Orthopedic, Physical and Functional Rehabilitation\nHospital of Tunis. The proposed intelligent system gives good results compared to the use of a standard joystick....
Behind firewalls, more and more cybersecurity attacks are specifically targeted to the very network where they are\ntaking place. This review proposes a comprehensive framework for addressing the challenge of characterising novel\ncomplex threats and relevant counter-measures. Two kinds of attacks are particularly representative of this issue:\nzero-day attacks that are not publicly disclosed and multi-step attacks that are built of several individual steps, some\nmalicious and some benign. Two main approaches are developed in the artificial intelligence field to track these\nattacks: statistics and machine learning. Statistical approaches include rule-based and outlier-detection-based\nsolutions. Machine learning includes the detection of behavioural anomalies and event sequence tracking.\nApplications of artificial intelligence cover the field of intrusion detection, which is typically performed online, and\nsecurity investigation, performed offline....
An accurate and efficient eye detector is essential for many computer vision applications. In this paper, we present an efficient\nmethod to evaluate the eye location from facial images. First, a group of candidate regions with regional extreme points is quickly\nproposed; then, a set of convolution neural networks (CNNs) is adopted to determine the most likely eye region and classify the\nregion as left or right eye; finally, the center of the eye is located with other CNNs. In the experiments using GI4E, BioID, and\nour datasets, our method attained a detection accuracy which is comparable to existing state-of-the-art methods; meanwhile, our\nmethod was faster and adaptable to variations of the images, including external light changes, facial occlusion, and changes in image\nmodality....
Loading....