Prefrontal cortical plasticity during learning of cognitive tasks

Training in working memory tasks is associated with lasting changes in prefrontal cortical activity. To assess the neural activity changes induced by training, we recorded single units, multi-unit activity (MUA) and local field potentials (LFP) with chronic electrode arrays implanted in the prefrontal cortex of two monkeys, throughout the period they were trained to perform cognitive tasks. Mastering different task phases was associated with distinct changes in neural activity, which included recruitment of larger numbers of neurons, increases or decreases of their firing rate, changes in the correlation structure between neurons, and redistribution of power across LFP frequency bands. In every training phase, changes induced by the actively learned task were also observed in a control task, which remained the same across the training period. Our results reveal how learning to perform cognitive tasks induces plasticity of prefrontal cortical activity, and how activity changes may generalize between tasks.

Aggregate-based Training Phase for ML-based Cardinality Estimation

Cardinality estimation is a fundamental task in database query processing and optimization. As shown in recent papers, machine learning (ML)-based approaches may deliver more accurate cardinality estimations than traditional approaches. However, a lot of training queries have to be executed during the model training phase to learn a data-dependent ML model making it very time-consuming. Many of those training or example queries use the same base data, have the same query structure, and only differ in their selective predicates. To speed up the model training phase, our core idea is to determine a predicate-independent pre-aggregation of the base data and to execute the example queries over this pre-aggregated data. Based on this idea, we present a specific aggregate-based training phase for ML-based cardinality estimation approaches in this paper. As we are going to show with different workloads in our evaluation, we are able to achieve an average speedup of 90 with our aggregate-based training phase and thus outperform indexes.

Fault Tolerance and Noise Immunity in Freespace Diffractive Optical Neural Networks

Free-space diffractive optical networks are a class of trainable optical media that are currently being explored as a novel hardware platform for neural engines. The training phase of such systems is usually performed in a computer and the learned weights are then transferred onto optical hardware ("ex-situ training"). Although this process of weight transfer has many practical advantages, it is often accompanied by performance degrading faults in the fabricated hardware. Being analog systems, these engines are also subject to performance degradation due to noises in the inputs and during optoelectronic conversion. Considering diffractive optical networks (DON) trained for image classification tasks on standard datasets, we numerically study the performance degradation arising out of weight faults and injected noises and methods to ameliorate these effects. Training regimens based on intentional fault and noise injection during the training phase are only found marginally successful at imparting fault tolerance or noise immunity. We propose an alternative training regimen using gradient based regularization terms in the training objective that are found to impart some degree of fault tolerance and noise immunity in comparison to injection based training regimen.

Implementation of the digital training concept in the basic flight training in the Serbian Military Academy

Introduction/purpose: The paper provides a review of recent research in the field of digital training applied in the Serbian Military Academy flight training. Flight training represents the foundation of successful education of military pilots. Its division is based on the environment and a phase of realization. The main part and the core of successful flight training is basic flight training. This training phase has experienced significant changes with the introduction of the Technically Advanced aircraft (TAA) which is characterized by a high degree of digitalization not only of the cockpit but also of other aircraft systems. Consequently, a different methodological approach to training is needed, including a digital training concept. The paper shows the achievements and certain solutions based on some elements of digital training concepts used in the basic flight training at the Serbian Military Academy. Methods: The scientific approach is used in the evaluation of aircraft cockpit digitalization and in the implementation of a new training concept in the basic flight training in the Serbian Military Academy. Results: Based on the methodological analysis used, the importance and the values of the digital training concept in basic flight training are shown. Conclusion: Although the digital training concept is not mandatory in the existing flight training model in the Serbian Military Academy, it proves to be a valuable asset. Its potential is significant and, to a certain extent, it can change the nature of basic flight training. Due to digital training, cadets can fly more safely and their flying skills are acquired faster. In accordance with new modern aircraft acquisition in the Serbian Air Force, every aspect of the digital training concept has to be carefully considered, especially in the basic flight training phase, including conversion to new aircraft types.

Writer Identification Based on Hand Writing using Artificial Neural Network

Humans are social beings who depend on social interaction. Social interaction that is often used is communication. Communication is one of the bridges to connect social relations between humans. Communication can be delivered in two ways, namely verbal or nonverbal. Handwriting is an example of nonverbal communication using paper and writing utensils. Each individual's writing has its own uniqueness so that handwriting often becomes the character or characteristic of the author. The handwriting pattern usually becomes a character for the writer so that people who recognize the writing will easily guess the ownership of the related handwriting. However, handwriting is often used by irresponsible people in the form of handwriting falsification. The acts of writing falcification often occur in the workplace or even in the field of education. This is one of the driving factors for creating a reliable system in tracking someone's handwriting based on their ownership. In this study, we will discuss the identification of a person's handwriting based on their ownership. The output of this research is in the form of ID from the author and accuracy in the form of percentage of system reliability in identifying. The results of this study are expected to have a good impact on all parties, in order to minimize plagiarism. Identification of handwriting to be built consists of two main processes, namely the training phase and the testing phase. At the training stage, the handwritten image is subjected to several processes, namely threshold, wavelet conversion, and then will be trained using the Backpropagation Artificial Neural Network. In the testing phase, the process is the same as in the training phase, but at the end of the process, a comparison will be made between the image data that has been stored during training with a comparison image. Backpropagation ANN can work optimally if it is trained using input data that has determined the size, learning rate, parameters, and the number of nodes on the network. It is expected that the offered method can work optimally so that it produces an accurate percentage in order to minimize handwriting falcification.

Automatic Phonemes Segmentation for Quran Verses Using Kaldi Toolkit

In this paper, automatic segmentation system was built using the Kaldi toolkit at phoneme level for Quran verses data set with a total speech corpus of (80 hours) and its corresponding text corpus respectively, with a size of 1100 recorded Quran verses of 100 non-Arab reciters. Initiated with the extraction of Mel Frequency Cepstral Coefficients MFCCs, the proceedings of the building of Language Model LM and Acoustic Model AM training phase continued until the Deep Neural Network DNN level by selecting 770 waves (70 reciters). The testing of the system was done using 220 waves (20 reciters), and concluded with the selection of the development data set which was 280 waves (10 reciters). Comparison was implemented between automatic and manual segmentation, and the results obtained for the test set was 99% and for the Development set was 99% with Time Delay Neural Networks TDNN based acoustic modelling.

Visual Perceptual Quality Assessment Based on Blind Machine Learning Techniques

This paper presents the construction of a new objective method for estimation of visual perceiving quality. The proposal provides an assessment of image quality without the need for a reference image or a specific distortion assumption. Two main processes have been used to build our models: The first one uses deep learning with a convolutional neural network process, without any preprocessing. The second objective visual quality is computed by pooling several image features extracted from different concepts: the natural scene statistic in the spatial domain, the gradient magnitude, the Laplacian of Gaussian, as well as the spectral and spatial entropies. The features extracted from the image file are used as the input of machine learning techniques to build the models that are used to estimate the visual quality level of any image. For the machine learning training phase, two main processes are proposed: The first proposed process consists of a direct learning using all the selected features in only one training phase, named direct learning blind visual quality assessment DLBQA. The second process is an indirect learning and consists of two training phases, named indirect learning blind visual quality assessment ILBQA. This second process includes an additional phase of construction of intermediary metrics used for the construction of the prediction model. The produced models are evaluated on many benchmarks image databases as TID2013, LIVE, and LIVE in the wild image quality challenge. The experimental results demonstrate that the proposed models produce the best visual perception quality prediction, compared to the state-of-the-art models. The proposed models have been implemented on an FPGA platform to demonstrate the feasibility of integrating the proposed solution on an image sensor.

TAE-Net: Task-Adaptive Embedding Network for Few-Shot Remote Sensing Scene Classification

Recently, approaches based on deep learning are quite prevalent in the area of remote sensing scene classification. Though significant success has been achieved, these approaches are still subject to an excess of parameters and extremely dependent on a large quantity of labeled data. In this study, few-shot learning is used for remote sensing scene classification tasks. The goal of few-shot learning is to recognize unseen scene categories given extremely limited labeled samples. For this purpose, a novel task-adaptive embedding network is proposed to facilitate few-shot scene classification of remote sensing images, referred to as TAE-Net. A feature encoder is first trained on the base set to learn embedding features of input images in the pre-training phase. Then in the meta-training phase, a new task-adaptive attention module is designed to yield the task-specific attention, which can adaptively select informative embedding features among the whole task. In the end, in the meta-testing phase, the query image derived from the novel set is predicted by the meta-trained model with limited support images. Extensive experiments are carried out on three public remote sensing scene datasets: UC Merced, WHU-RS19, and NWPU-RESISC45. The experimental results illustrate that our proposed TAE-Net achieves new state-of-the-art performance for few-shot remote sensing scene classification.

Time-lapse seismic data inversion for estimating reservoir parametersusing deep learning

Geological carbon sequestration involves the injection of captured carbon dioxide (CO2) into sub-surface formations for long-term storage. The movement and fate of the injected CO2 plume is ofgreat concern to regulators as monitoring helps to identify potential leakage zones and determinesthe possibility of safe long-term storage. To address this concern, we design a deep learning frame-work for carbon dioxide (CO2) saturation monitoring to determine the geological controls on thestorage of the injected CO2. We use different combinations of porosities and permeabilities for agiven reservoir to generate saturation and velocity models. We train the deep learning model with afew time-lapse seismic images and their corresponding changes in saturation values for a particular CO2 injection site. The deep learning model learns the mapping from the change in the time-lapseseismic response to the change in CO2 saturation during the training phase. We then apply thetrained model to data sets comprising different time-lapse seismic image slices (corresponding todifferent time instances) generated using different porosity and permeability distributions that arenot part of the training to estimate the CO2 saturation values along with the plume extent. Theproposed algorithm provides a deep learning assisted framework for the direct estimation of CO2 saturation values and plume migration in heterogeneous formations using the time-lapse seismicdata. The proposed method improves the efficiency of time-lapse inversion by streamlining thelarge number of intermediate steps in the conventional time-lapse inversion workflow. This method also helps to incorporate the geological uncertainty for a given reservoir by accounting for the statis-tical distribution of porosity and permeability during the training phase. Tests on different examplesverify the effectiveness of the proposed approach

Veränderung des NPC bei Personen mit Konvergenzinsuffizienz nach I.F.S.-Visualtraining

Purpose. The aim of this study was to evaluate the efficacy of the I.F.S. visual therapy (convergence training) developed by Bruce Evans in a randomized-controlled setting. The analysis of the changes of the near point of convergence (NPC) was the main target of the study. Material and Methods. 20 subjects (39.0 ± 15.32 years) with convergence insufficiency and an NPC > 10 cm underwent 4 weeks of visual therapy in a randomised-controlled singleblind setting. The verum group followed the I.F.S. exercises, while the control group performed a placebo therapy. Before and after the training phase, the NPC and the positive fusional vergence (PFV) were measured in all subjects and the Sheard and Mallett criteria were also assessed. The CISS questionnaire was used to record and quantify subjective symptoms. Results. The statistical analysis (α = 0.05) shows that the NPC of the verum group improved significantly compared to the control group (p = 0.0008) and within the verum group only (p = 0.0002). The positive effect of the I.F.S. exercises is confirmed by the also significant improvement of the PFV. Conclusion. The results as well as the practical experience with the exercises indicate that the I.F.S. visual therapy proves to be effective for individuals with a convergence insufficiency and an NPC > 10 cm in improving the near point of convergence. Keywords. convergence insufficiency, near point of convergence, NPC, visual therapy, convergence training, I.F.S. exercises