Search results for: convolutional long short-term memory

Authors: Manveen Singh

Abstract:

Since the very advent of law enforcement, eyewitness testimony has played a pivotal role in identifying, arresting and convicting suspects. Reliant heavily on the accuracy of human memory, nothing seems to carry more weight with the judiciary than the testimony of an actual witness. The acceptance of eyewitness testimony as a substantive piece of evidence lies embedded in the assumption that the human mind is adept at recording and storing events. Research though, has proven otherwise. Having carried out extensive study in the field of eyewitness testimony for the past 40 years, psychologists have concluded that human memory is fragile and needs to be treated carefully. The question that arises then, is how reliable is eyewitness testimony? The credibility of eyewitness testimony, simply put, depends on several factors leaving it reliable at times while not so much at others. This is further substantiated by the fact that as per scientific research, over 75 percent of all eyewitness testimonies may stand in error; quite a few of these cases resulting in life sentences. Although the advancement of scientific techniques, especially DNA testing, helped overturn many of these eyewitness testimony-based convictions, yet eyewitness identifications continue to form the backbone of most police investigations and courtroom decisions till date. What then is the solution to this long standing concern regarding the accuracy of eyewitness accounts? The present paper shall analyze the linkage between human memory and eyewitness identification as well as look at the various factors governing the credibility of eyewitness testimonies. Furthermore, it shall elaborate upon some best practices developed over the years to help reduce mistaken identifications. Thus, in the process, trace out the changing landscape of eyewitness testimony amidst the evolution of DNA and trace evidence.

Keywords: DNA, eyewitness, identification, testimony, evidence

Procedia PDF Downloads 308

Authors: Anchana Julsiri, Seree Chadcham

Abstract:

The present study determined the effects of listening to pleasant Thai classical music on increasing working memory in elderly. Thai classical music without lyrics that made participants feel fun and aroused was used in the experiment for 3.19-5.40 minutes. The accuracy scores of Counting Span Task (CST), upper alpha ERD%, and theta ERS% were used to assess working memory of participants both before and after listening to pleasant Thai classical music. The results showed that the accuracy scores of CST and upper alpha ERD% in the frontal area of participants after listening to Thai classical music were significantly higher than before listening to Thai classical music (p < .05). Theta ERS% in the fronto-parietal network of participants after listening to Thai classical music was significantly lower than before listening to Thai classical music (p < .05).

Keywords: brain wave, elderly, pleasant Thai classical music, working memory

Procedia PDF Downloads 429

Authors: Gizem Eser Erdek

Abstract:

This study is research on predicting the remaining life of industrial cutting tools used in the industrial production process with deep learning methods. When the life of cutting tools decreases, they cause destruction to the raw material they are processing. This study it is aimed to predict the remaining life of the cutting tool based on the damage caused by the cutting tools to the raw material. For this, hole photos were collected from the hole-drilling machine for 8 months. Photos were labeled in 5 classes according to hole quality. In this way, the problem was transformed into a classification problem. Using the prepared data set, a model was created with convolutional neural networks, which is a deep learning method. In addition, VGGNet and ResNet architectures, which have been successful in the literature, have been tested on the data set. A hybrid model using convolutional neural networks and support vector machines is also used for comparison. When all models are compared, it has been determined that the model in which convolutional neural networks are used gives successful results of a %74 accuracy rate. In the preliminary studies, the data set was arranged to include only the best and worst classes, and the study gave ~93% accuracy when the binary classification model was applied. The results of this study showed that the remaining life of the cutting tools could be predicted by deep learning methods based on the damage to the raw material. Experiments have proven that deep learning methods can be used as an alternative for cutting tool life estimation.

Keywords: classification, convolutional neural network, deep learning, remaining life of industrial cutting tools, ResNet, support vector machine, VggNet

Procedia PDF Downloads 44

Authors: Bingquan Shen

Abstract:

Convolutional Neural Networks (CNN) have demonstrated their effectiveness in synthesizing 3D views of object instances at various viewpoints. Given the problem where one have limited viewpoints of a particular object for classification, we present a pose normalization architecture to transform the object to existing viewpoints in the training dataset before classification to yield better classification performance. We have demonstrated that this Pose Normalization Network (PNN) can capture the style of the target object and is able to re-render it to a desired viewpoint. Moreover, we have shown that the PNN improves the classification result for the 3D chairs dataset and ShapeNet airplanes dataset when given only images at limited viewpoint, as compared to a CNN baseline.

Keywords: convolutional neural networks, object classification, pose normalization, viewpoint invariant

Procedia PDF Downloads 306

Authors: Joseph George, Anne Kotteswara Roa

Abstract:

Skin disease is one of the most common and popular kinds of health issues faced by people nowadays. Skin cancer (SC) is one among them, and its detection relies on the skin biopsy outputs and the expertise of the doctors, but it consumes more time and some inaccurate results. At the early stage, skin cancer detection is a challenging task, and it easily spreads to the whole body and leads to an increase in the mortality rate. Skin cancer is curable when it is detected at an early stage. In order to classify correct and accurate skin cancer, the critical task is skin cancer identification and classification, and it is more based on the cancer disease features such as shape, size, color, symmetry and etc. More similar characteristics are present in many skin diseases; hence it makes it a challenging issue to select important features from a skin cancer dataset images. Hence, the skin cancer diagnostic accuracy is improved by requiring an automated skin cancer detection and classification framework; thereby, the human expert’s scarcity is handled. Recently, the deep learning techniques like Convolutional neural network (CNN), Deep belief neural network (DBN), Artificial neural network (ANN), Recurrent neural network (RNN), and Long and short term memory (LSTM) have been widely used for the identification and classification of skin cancers. This survey reviews different DL techniques for skin cancer identification and classification. The performance metrics such as precision, recall, accuracy, sensitivity, specificity, and F-measures are used to evaluate the effectiveness of SC identification using DL techniques. By using these DL techniques, the classification accuracy increases along with the mitigation of computational complexities and time consumption.

Keywords: skin cancer, deep learning, performance measures, accuracy, datasets

Procedia PDF Downloads 95

Authors: Mackenzie Leake, Liyu Xia, Kamil Rocki, Wayne Imaino

Abstract:

In the Hierarchical Temporal Memory (HTM) paradigm the effect of overlap between inputs on the activation of columns in the spatial pooler is studied. Numerical results suggest that similar inputs are represented by similar sets of columns and dissimilar inputs are represented by dissimilar sets of columns. It is shown that the spatial pooler produces these results under certain conditions for the connectivity and proximal thresholds. Following the discussion of the initialization of parameters for the thresholds, corresponding qualitative arguments about the learning dynamics of the spatial pooler are discussed.

Keywords: hierarchical temporal memory, HTM, learning algorithms, machine learning, spatial pooler

Procedia PDF Downloads 314

Authors: Toledo-Fernandez Aldebaran

Abstract:

Introduction: One of the least explored cognitive functions in relation with depression is the one related to musical stimuli. Music perception and memory can become impaired as well. The term amusia is used to define a type of agnosia caused by damage to basic processes that creates a general inability to perceive music. Therefore, the main objective is to explore performance-based and self-report deficits in music perception and memory on people with major depressive disorder (MDD). Method: Data was collected through April-October 2021 recruiting people who met the eligibility criteria and using the Montreal Battery of Evaluation of Amusia (MBEA) to evaluate performance-based music perception and memory, along with the module for depression of the Mini International Neuropsychiatric Interview, and the Amusic Dysfunction Inventory (ADI) which evaluates the participants’ self-report concerning their abilities in music perception. Results: 64 participants were evaluated. The main study, referring to analyzing the differences between people with MDD and the control group, only showed one statistical difference on the Interval subtest of the MBEA. No difference was found in the dimensions assessed by the ADI. Conclusion: Deficits in interval perception can be explained by mental fatigue, to which people with depression are more vulnerable, rather than by specific deficits in musical perception and memory associated with depressive disorder. Additionally, significant associations were found between musical deficits as observed by performance-based evidence and music dysfunction according to self-report, which could suggest that some people with depression are capable of detecting these deficits in themselves.

Keywords: depression, amusia, music, perception, memory

Procedia PDF Downloads 30

Authors: Pedro Henrique Alves Martins, Paulo Guilherme Ferreira De Siqueira, Franco De Castro Bubani, Maria Teresa Paulino Aguilar, Paulo Roberto Cetlin

Abstract:

Cu-base shape-memory alloys (CuZnAl, CuAlNi, CuAlBe, etc.) are promising engineering materials for several unconventional devices, such as sensors, actuators, and mechanical vibration dampers. Brittleness is one of the factors that limit the commercial use of these alloys, as it makes thermomechanical processing difficult. In this work, a method for the hot extrusion of a 75.50% Cu, 16,74% Zn, 7,76% Al (weight %) alloy is presented. The effects of the thermomechanical processing in the microstructure and the pseudoelastic behavior of the alloy are assessed by optical metallography, compression and hardness tests. Results show that hot extrusion is a suitable method to obtain severe cross-section reductions in the CuZnAl shape-memory alloy studied. The alloy maintained its pseudoelastic effect after the extrusion and the modifications in the mechanical behavior caused by precipitation during hot extrusion can be minimized by a suitable precipitate dissolution heat treatment.

Keywords: hot extrusion, pseudoelastic, shape-memory alloy, thermomechanical processing

Procedia PDF Downloads 346

Authors: Jiahao Cui

Abstract:

In the past centuries, due to climate change and intense whaling, the global whale population has dramatically declined. Among the various whale species, the fin whale experienced the most drastic drop in number due to its popularity in whaling. Under this background, identifying fin whale calls could be immensely beneficial to the preservation of the species. This paper uses feature extraction to process the input audio signal, then a network based on AlexNet and three networks based on the ResNet model was constructed to classify fin whale calls. A mixture of the DOSITS database and the Watkins database was used during training. The results demonstrate that a modified ResNet network has the best performance considering precision and network complexity.

Keywords: convolutional neural network, ResNet, AlexNet, fin whale preservation, feature extraction

Procedia PDF Downloads 88

Authors: Kamyar Fakhr, Roozbeh Salmani

Abstract:

Alphonse Bertillon developed the first biometric security system in the 1800s. Today, many governments and giant companies are considering or have procured biometrically enabled security schemes. Iris is a kaleidoscope of patterns and colors. Each individual holds a set of irises more unique than their thumbprint. Every single day, giant companies like Google and Apple are experimenting with reliable biometric systems. Now, after almost 200 years of improvements, face ID does not work with masks, it gives access to fake 3D images, and there is no global usage of biometric recognition systems as national identity (ID) card. The goal of this paper is to demonstrate the advantages of iris recognition overall biometric recognition systems. It make two extensions: first, we illustrate how a very large amount of internet fraud and cyber abuse is happening due to bugs in face recognition systems and in a very large dataset of 3.4M people; second, we discuss how establishing a secure global network of iris recognition devices connected to authoritative convolutional neural networks could be the safest solution to this dilemma. Another aim of this study is to provide a system that will prevent system infiltration caused by cyber-attacks and will block all wireframes to the data until the main user ceases the procedure.

Keywords: biometric system, convolutional neural network, cyber-attack, secure

Procedia PDF Downloads 189

Authors: Mo-Se Lee, Cheol-Hwi Ahn, Kee-Young Kwahk, Hyunchul Ahn

Abstract:

Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN (Convolutional Neural Network), which is known as effective solution for recognizing and classifying images, has been popularly applied to classification and prediction problems in various fields. In this study, we try to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. In specific, we propose to apply CNN as the binary classifier that predicts stock market direction (up or down) by using a graph as its input. That is, our proposal is to build a machine learning algorithm that mimics a person who looks at the graph and predicts whether the trend will go up or down. Our proposed model consists of four steps. In the first step, it divides the dataset into 5 days, 10 days, 15 days, and 20 days. And then, it creates graphs for each interval in step 2. In the next step, CNN classifiers are trained using the graphs generated in the previous step. In step 4, it optimizes the hyper parameters of the trained model by using the validation dataset. To validate our model, we will apply it to the prediction of KOSPI200 for 1,986 days in eight years (from 2009 to 2016). The experimental dataset will include 14 technical indicators such as CCI, Momentum, ROC and daily closing price of KOSPI200 of Korean stock market.

Keywords: convolutional neural network, deep learning, Korean stock market, stock market prediction

Procedia PDF Downloads 406

Authors: Zhongmin Wang, Wudong Fan, Hengshan Zhang, Yimin Zhou

Abstract:

In data-driven prognostic methods, the prediction accuracy of the estimation for remaining useful life of bearings mainly depends on the performance of health indicators, which are usually fused some statistical features extracted from vibrating signals. However, the existing health indicators have the following two drawbacks: (1) The differnet ranges of the statistical features have the different contributions to construct the health indicators, the expert knowledge is required to extract the features. (2) When convolutional neural networks are utilized to tackle time-frequency features of signals, the time-series of signals are not considered. To overcome these drawbacks, in this study, the method combining convolutional neural network with gated recurrent unit is proposed to extract the time-frequency image features. The extracted features are utilized to construct health indicator and predict remaining useful life of bearings. First, original signals are converted into time-frequency images by using continuous wavelet transform so as to form the original feature sets. Second, with convolutional and pooling layers of convolutional neural networks, the most sensitive features of time-frequency images are selected from the original feature sets. Finally, these selected features are fed into the gated recurrent unit to construct the health indicator. The results state that the proposed method shows the enhance performance than the related studies which have used the same bearing dataset provided by PRONOSTIA.

Keywords: continuous wavelet transform, convolution neural net-work, gated recurrent unit, health indicators, remaining useful life

Procedia PDF Downloads 101

Authors: Rana Zeina

Abstract:

Objective: The association between Sensory problems and cognitive abilities has been studied in individuals with Autism Spectrum Disorders (ASDs). In this study, we used a neuropsychological test to evaluate memory and attention in ASDs children with sensory problems compared to the ASDs children without sensory problems. Methods: Four visual memory tests of Cambridge Neuropsychological Test Automated Battery (CANTAB) including Big/Little Circle (BLC), Simple Reaction Time (SRT), Intra/Extra Dimensional Set Shift (IED), Spatial Recognition Memory (SRM), were administered to 14 ASDs children with sensory problems compared to 13 ASDs without sensory problems aged 3 to 12 with IQ of above 70. Results: ASDs Individuals with sensory problems performed worse than the ASDs group without sensory problems on comprehension, learning, reversal and simple reaction time tasks, and no significant difference between the two groups was recorded in terms of the visual memory and visual comprehension tasks. Conclusion: The findings of this study suggest that ASDs children with sensory problems are facing deficits in learning, comprehension, reversal, and speed of response to stimuli.

Keywords: visual memory, attention, autism spectrum disorders, CANTAB eclipse

Procedia PDF Downloads 424

Authors: Priyanka Bhatnagar, Malkeshkumar Patel, Joondong Kim, Joonpyo Hong

Abstract:

Artificial skin and sensory memory platforms are produced using a flexible, transparent photovoltaic (TPV) device. The TPV device is composed of a metal oxide heterojunction (nZnO/p-NiO) and transmits visible light (> 50%) while producing substantial electric power (0.5 V and 200 μA cm-2 ). This TPV device is a transparent energy interface that can be used to detect signals and propagate information without an external energy supply. The TPV artificial skin offers a temperature detection range (0 C75 C) that is wider than that of natural skin (5 C48 °C) due to the temperature-sensitive pyrocurrent from the ZnO layer. Moreover, the TPV thermoreceptor offers sensory memory of extreme thermal stimuli. Much like natural skin, artificial skin uses the nociceptor mechanism to protect tissue from harmful damage via signal amplification (hyperalgesia) and early adaption (allodynia). This demonstrates the many features of TPV artificial skin, which can sense and transmit signals and memorize information under self-operation mode. This transparent photovoltaic skin can provide sustainable energy for use in human electronics.

Keywords: transparent, photovoltaics, thermal memory, artificial skin, thermoreceptor

Procedia PDF Downloads 84

Authors: Farzaneh Badinlou, Reza Kormi-Nouri, Monika Knopf

Abstract:

Studies of adults' episodic memory have found that enacted encoding not only improve recall performance but also retrieve faster during the recall period. The current study focused on exploring the temporal progression of different encoding conditions in younger and older school children. 204 students from two age group of 8 and 14 participated in this study. During the study phase, we studied action encoding in two forms; participants performed the phrases by themselves (SPT), and observed the performance of the experimenter (EPT), which were compared with verbal encoding; participants listened to verbal action phrases (VT). At test phase, we used immediate and delayed free recall tests. We observed significant differences in memory performance as function of age group, and encoding conditions in both immediate and delayed free recall tests. Moreover, temporal progression of recall was faster in older children when compared with younger ones. The interaction of age-group and encoding condition was only significant in delayed recall displaying that younger children performed better in EPT whereas older children outperformed in SPT. It was proposed that enactment effect in form of SPT enhances item-specific processing, whereas EPT improves relational information processing and this differential processes are responsible for the results achieved in younger and older children. The role of memory strategies and information processing methods in younger and older children were considered in this study. Moreover, the temporal progression of recall was faster in action encoding in the form of SPT and EPT compared with verbal encoding in both immediate and delayed free recall and size of enactment effect was constantly increased throughout the recall period. The results of the present study provide further evidence that the action memory is explained with an emphasis on the notion of information processing and strategic views. These results also reveal the temporal progression of recall as a new dimension of episodic memory in children.

Keywords: action memory, enactment effect, episodic memory, school-aged children, temporal progression

Procedia PDF Downloads 246

Authors: Vladimir A. Vinnikov

Abstract:

The paper proposes a model of an inhomogeneous rock mass with initially random distribution of microcracks on mineral grain boundaries. It describes the behavior of cracks in a medium under the effect of thermal field, the medium heated instantaneously to a predetermined temperature. Crack growth occurs according to the concept of fracture mechanics provided that the stress intensity factor K exceeds the critical value of Kc. The modeling of thermally induced acoustic emission memory effects is based on the assumption that every event of crack nucleation or crack growth caused by heating is accompanied by a single acoustic emission event. Parameters of the thermally induced acoustic emission memory effect produced by cyclic heating and cooling (with the temperature amplitude increasing from cycle to cycle) were calculated for several rock texture types (massive, banded, and disseminated). The study substantiates the adaptation of the proposed model to humidity interference with the thermally induced acoustic emission memory effect. The influence of humidity on the thermally induced acoustic emission memory effect in quasi-homogeneous and banded rocks is estimated. It is shown that such modeling allows the structure and texture of rocks to be taken into account and the influence of interference factors on the distinctness of the thermally induced acoustic emission memory effect to be estimated. The numerical modeling can be used to obtain information about the thermal impacts on rocks in the past and determine the degree of rock disturbance by means of non-destructive testing.

Keywords: degree of rock disturbance, non-destructive testing, thermally induced acoustic emission memory effects, structure and texture of rocks

Procedia PDF Downloads 235

Authors: Gong Zhilin, Jing Yang, Jian Yin

Abstract:

The purpose of this paper is to develop a novel system for credit card fraud detection based on sequential modeling of data using hybrid deep learning models. The projected model encapsulates five major phases are pre-processing, imbalance-data handling, feature extraction, optimal feature selection, and fraud detection with an ensemble classifier. The collected raw data (input) is pre-processed to enhance the quality of the data through alleviation of the missing data, noisy data as well as null values. The pre-processed data are class imbalanced in nature, and therefore they are handled effectively with the K-means clustering-based SMOTE model. From the balanced class data, the most relevant features like improved Principal Component Analysis (PCA), statistical features (mean, median, standard deviation) and higher-order statistical features (skewness and kurtosis). Among the extracted features, the most optimal features are selected with the Self-improved Arithmetic Optimization Algorithm (SI-AOA). This SI-AOA model is the conceptual improvement of the standard Arithmetic Optimization Algorithm. The deep learning models like Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), and optimized Quantum Deep Neural Network (QDNN). The LSTM and CNN are trained with the extracted optimal features. The outcomes from LSTM and CNN will enter as input to optimized QDNN that provides the final detection outcome. Since the QDNN is the ultimate detector, its weight function is fine-tuned with the Self-improved Arithmetic Optimization Algorithm (SI-AOA).

Keywords: credit card, data mining, fraud detection, money transactions

Procedia PDF Downloads 102

Authors: Ercan Eker, Muhammer Karaman, Akif Aslan, Hakan Tanrikuluoglu

Abstract:

Deficiency of institutional memory and knowledge management can result in information security breaches, loss of prestige and trustworthiness and the worst the loss of know-how and institutional knowledge. Traditional learning management within organizations is generally handled by personal efforts. That kind of struggle mostly depends on personal desire, motivation and institutional belonging. Even if an organization has highly motivated employees at a certain time, the institutional knowledge and memory life cycle will generally remain limited to these employees’ spending time in this organization. Having a learning management system in an organization can sustain the institutional memory, knowledge and know-how in the organization. Learning management systems are much more needed especially in public organizations where the job rotation is frequently seen and managers are appointed periodically. However, a learning management system should not be seen as an organizations’ website. It is a more comprehensive, interactive and user-friendly knowledge management tool for organizations. In this study, the importance of using learning management systems in the process of emerging tacit knowledge is underlined.

Keywords: knowledge management, learning management systems, tacit knowledge, institutional memory

Procedia PDF Downloads 344

Authors: Laura M. F. Bertens

Abstract:

The houses we inhabit are important containers of memory. As homes, they take on meaning for those who live inside, and memories of family life become intimately tied up with rooms, windows, and gardens. Each new family creates a new layer of meaning, resulting in a palimpsest of family memory. These houses function quite literally as memory palaces, as a walk through a childhood home will show; each room conjures up images of past events. Over time, these personal memories become woven together with the cultural memory of countries and generations. The importance of the home is a central theme in art, and several contemporary artists have a special interest in the relationship between memory and the home. This paper analyses three case studies in order to get a deeper understanding of the ways in which the home functions and feels like a memory palace, both on an individual and on a collective, cultural level. Close reading of the artworks is performed on the theoretical intersection between Art History and Cultural Memory Studies. The first case study concerns works from the exhibition Mnemosyne by the artist duo Anne and Patrick Poirier. These works combine interests in architecture, archaeology, and psychology. Models of cities and fantastical architectural designs resemble physical structures (such as the brain), architectural metaphors used in representing the concept of memory (such as the memory palace), and archaeological remains, essential to our shared cultural memories. Secondly, works by Do Ho Suh will help us understand the relationship between the home and memory on a far more personal level; outlines of rooms from his former homes, made of colourful, transparent fabric and combined into new structures, provide an insight into the way these spaces retain individual memories. The spaces have been emptied out, and only the husks remain. Although the remnants of walls, light switches, doors, electricity outlets, etc. are standard, mass-produced elements found in many homes and devoid of inherent meaning, together they remind us of the emotional significance attached to the muscle memory of spaces we once inhabited. The third case study concerns an exhibition in a house put up for sale on the Dutch real estate website Funda. The house was built in 1933 by a Jewish family fleeing from Germany, and the father and son were later deported and killed. The artists Anne van As and CA Wertheim have used the history and memories of the house as a starting point for an exhibition called (T)huis, a combination of the Dutch words for home and house. This case study illustrates the way houses become containers of memories; each new family ‘resets’ the meaning of a house, but traces of earlier memories remain. The exhibition allows us to explore the transition of individual memories into shared cultural memory, in this case of WWII. Taken together, the analyses provide a deeper understanding of different facets of the relationship between the home and memory, both individual and collective, and the ways in which art can represent these.

Keywords: Anne and Patrick Poirier, cultural memory, Do Ho Suh, home, memory palace

Procedia PDF Downloads 129

Authors: Girish J. Showkatramani, Arthi M. Krishna, Sashi Nareddi, Naresh Nula, Aaron Pepe, Glen Brown, Greg Gabel, Chris Doninger

Abstract:

Trademark examination and approval is a complex process that involves analysis and review of the design components of the marks such as the visual representation as well as the textual data associated with marks such as marks' description. Currently, the process of identifying marks with similar visual representation is done manually in United States Patent and Trademark Office (USPTO) and takes a considerable amount of time. Moreover, the accuracy of these searches depends heavily on the experts determining the trademark design codes used to catalog the visual design codes in the mark. In this study, we explore several methods to automate trademark design code classification. Based on recent successes of convolutional neural networks in image classification, we have used several different convolutional neural networks such as Google’s Inception v3, Inception-ResNet-v2, and Xception net. The study also looks into other techniques to augment the results from CNNs such as using Open Source Computer Vision Library (OpenCV) to pre-process the images. This paper reports the results of the various models trained on year of annotated trademark images.

Keywords: trademark design code, convolutional neural networks, trademark image classification, trademark image search, Inception-ResNet-v2

Procedia PDF Downloads 198

Authors: Anthony D. Rhodes, Manan Goel

Abstract:

We provide a high fidelity deep learning algorithm (HyperSeg) for interactive video segmentation tasks using a dense convolutional network with context-aware skip connections and compressed, 'hypercolumn' image features combined with a convolutional tessellation procedure. In order to maintain high output fidelity, our model crucially processes and renders all image features in high resolution, without utilizing downsampling or pooling procedures. We maintain this consistent, high grade fidelity efficiently in our model chiefly through two means: (1) we use a statistically-principled, tensor decomposition procedure to modulate the number of hypercolumn features and (2) we render these features in their native resolution using a convolutional tessellation technique. For improved pixel-level segmentation results, we introduce a boundary loss function; for improved temporal coherence in video data, we include temporal image information in our model. Through experiments, we demonstrate the improved accuracy of our model against baseline models for interactive segmentation tasks using high resolution video data. We also introduce a benchmark video segmentation dataset, the VFX Segmentation Dataset, which contains over 27,046 high resolution video frames, including green screen and various composited scenes with corresponding, hand-crafted, pixel-level segmentations. Our work presents a improves state of the art segmentation fidelity with high resolution data and can be used across a broad range of application domains, including VFX pipelines and medical imaging disciplines.

Keywords: computer vision, object segmentation, interactive segmentation, model compression

Procedia PDF Downloads 96

Authors: Carlos Tarmeño Noriega, Igor Aguilar Alonso

Abstract:

Fungal diseases are common in vineyards; they cause a decrease in the quality of the products that can be sold, generating distrust of the customer towards the seller when buying fruit. Currently, technology allows the classification of fruits according to their characteristics thanks to artificial intelligence. This study proposes the implementation of a control system that allows the identification of the main fungal diseases present in the Italia grape, making use of a convolutional neural network (CNN), OpenCV, and TensorFlow. The methodology used was based on a collection of 20 articles referring to the proposed research on quality control, classification, and recognition of fruits through artificial vision techniques.

Keywords: computer vision, convolutional neural networks, quality control, fruit market, OpenCV, TensorFlow

Procedia PDF Downloads 47

Authors: Mohamad Bagher Heidari, Hefzollah Mohammadian

Abstract:

New discoveries in materials on the nanometer-length scale are expected to play an important role in addressing ongoing and future challenges in the field of communication. Devices and systems for ultra-high speed short and long range communication links, portable and power efficient computing devices, high-density memory and logics, ultra-fast interconnects, and autonomous and robust energy scavenging devices for accessing ambient intelligence and needed information will critically depend on the success of next-generation emerging nonmaterials and devices. This article presents some exciting recent developments in nonmaterials that have the potential to play a critical role in the development and transformation of future intelligent communication and telecom networks in the gas refinery. The industry is benefiting from nanotechnology advances with numerous applications including those in smarter sensors, logic elements, computer chips, memory storage devices, optoelectronics.

Keywords: nonmaterial, intelligent communication, nanoscale, nanophotonic, telecom

Procedia PDF Downloads 291

Authors: Varun Agarwal

Abstract:

Introduction: With the advent of personalized medicine, histopathological review of whole slide images (WSIs) for cancer diagnosis presents an exceedingly time-consuming, complex task. Specifically, detecting metastatic regions in WSIs of sentinel lymph node biopsies necessitates a full-scanned, holistic evaluation of the image. Thus, digital pathology, low-level image manipulation algorithms, and machine learning provide significant advancements in improving the efficiency and accuracy of WSI analysis. Using Camelyon16 data, this paper proposes a deep learning pipeline to automate and ameliorate breast cancer metastasis localization and WSI classification. Methodology: The model broadly follows five stages -region of interest detection, WSI partitioning into image tiles, convolutional neural network (CNN) image-segment classifications, probabilistic mapping of tumor localizations, and further processing for whole WSI classification. Transfer learning is applied to the task, with the implementation of Inception-ResNetV2 - an effective CNN classifier that uses residual connections to enhance feature representation, adding convolved outputs in the inception unit to the proceeding input data. Moreover, in order to augment the performance of the transfer learning CNN, a stack of convolutional denoising autoencoders (CDAE) is applied to produce embeddings that enrich image representation. Through a saliency-detection algorithm, visual training segments are generated, which are then processed through a denoising autoencoder -primarily consisting of convolutional, leaky rectified linear unit, and batch normalization layers- and subsequently a contrast-normalization function. A spatial pyramid pooling algorithm extracts the key features from the processed image, creating a viable feature map for the CNN that minimizes spatial resolution and noise. Results and Conclusion: The simplified and effective architecture of the fine-tuned transfer learning Inception-ResNetV2 network enhanced with the CDAE stack yields state of the art performance in WSI classification and tumor localization, achieving AUC scores of 0.947 and 0.753, respectively. The convolutional feature retention and compilation with the residual connections to inception units synergized with the input denoising algorithm enable the pipeline to serve as an effective, efficient tool in the histopathological review of WSIs.

Keywords: breast cancer, convolutional neural networks, metastasis mapping, whole slide images

Procedia PDF Downloads 105

Authors: Vikrant Gupta, Amrit Goswami

Abstract:

The fixed income market forms the basis of the modern financial market. All other assets in financial markets derive their value from the bond market. Owing to its over-the-counter nature, corporate bonds have relatively less data publicly available and thus is researched upon far less compared to Equities. Bond price prediction is a complex financial time series forecasting problem and is considered very crucial in the domain of finance. The bond prices are highly volatile and full of noise which makes it very difficult for traditional statistical time-series models to capture the complexity in series patterns which leads to inefficient forecasts. To overcome the inefficiencies of statistical models, various machine learning techniques were initially used in the literature for more accurate forecasting of time-series. However, simple machine learning methods such as linear regression, support vectors, random forests fail to provide efficient results when tested on highly complex sequences such as stock prices and bond prices. hence to capture these intricate sequence patterns, various deep learning-based methodologies have been discussed in the literature. In this study, a recurrent neural network-based deep learning model using long short term networks for prediction of corporate bond prices has been discussed. Long Short Term networks (LSTM) have been widely used in the literature for various sequence learning tasks in various domains such as machine translation, speech recognition, etc. In recent years, various studies have discussed the effectiveness of LSTMs in forecasting complex time-series sequences and have shown promising results when compared to other methodologies. LSTMs are a special kind of recurrent neural networks which are capable of learning long term dependencies due to its memory function which traditional neural networks fail to capture. In this study, a simple LSTM, Stacked LSTM and a Masked LSTM based model has been discussed with respect to varying input sequences (three days, seven days and 14 days). In order to facilitate faster learning and to gradually decompose the complexity of bond price sequence, an Empirical Mode Decomposition (EMD) has been used, which has resulted in accuracy improvement of the standalone LSTM model. With a variety of Technical Indicators and EMD decomposed time series, Masked LSTM outperformed the other two counterparts in terms of prediction accuracy. To benchmark the proposed model, the results have been compared with traditional time series models (ARIMA), shallow neural networks and above discussed three different LSTM models. In summary, our results show that the use of LSTM models provide more accurate results and should be explored more within the asset management industry.

Keywords: bond prices, long short-term memory, time series forecasting, empirical mode decomposition

Procedia PDF Downloads 105

Authors: Navonil Hazra

Abstract:

The paper discusses how and why aphasia can be understood as the language of nonverbal communication in a musical performance. The elements that are required to classify it as a nonverbal language. Since music is regarded as a nonverbal medium that cannot be engaged in any language, it is concerned about how aphasia might be called the language of nonverbalization. The paper also talks about how it portrays the magnificence of a performance, and how it expresses its likings or dislikes. Regarding the reasons for aphasia, the paper talks about the seizure factor and elucidates on seizure subjects as well. Furthermore, it discusses collective seizures and individual seizures. It also tries to consider aphasia as a-posteriori rather than a-priori looking at it from the lens of ‘Pure Reason’. Along with aphasia, the paper tries to make a critique of silence and the possibilities of looking at silence differently, also looking at the ontology of silence and sound. This paper also critically examines silence and the significance of gestures in performance. It also investigates whether gestures are accompanied by silence, establishing the notion of agential silence. This paper also talks about the place and role of memory in the formulation and analysis of a performance, as well as the plaguing and reclamation of memory, how memory alters the linear course of time and taunts us to look for alternative models of temporalities. This paper discusses the concept of 'auditory labour', with active and passive listening.

Keywords: aphasia, gestures, memory, silence

Procedia PDF Downloads 6

Authors: Tanvi P. Patel, Warish D. Patel

Abstract:

Association rule mining forms the core of data mining and it is termed as one of the well-known methodologies of data mining. Objectives of mining is to find interesting correlations, frequent patterns, associations or casual structures among sets of items in the transaction databases or other data repositories. Hence, association rule mining is imperative to mine patterns and then generate rules from these obtained patterns. For efficient targeted query processing, finding frequent patterns and itemset mining, there is an efficient way to generate an itemset tree structure named Memory Efficient Itemset Tree. Memory efficient IT is efficient for storing itemsets, but takes more time as compare to traditional IT. The proposed strategy generates maximal frequent itemsets from memory efficient itemset tree by using levelwise pruning. For that firstly pre-pruning of items based on minimum support count is carried out followed by itemset tree reconstruction. By having maximal frequent itemsets, less number of patterns are generated as well as tree size is also reduced as compared to MEIT. Therefore, an enhanced approach of memory efficient IT proposed here, helps to optimize main memory overhead as well as reduce processing time.

Keywords: association rule mining, itemset mining, itemset tree, meit, maximal frequent pattern

Procedia PDF Downloads 343

Authors: Diana Grguric, Robert Svetlacic, Vladimir Simovic

Abstract:

Scientific research analytically explores database playlists by studying the memory culture through Croatian popular radio music. The research is based on the scientific analysis of databases developed on the basis of the playlist of ten Croatian radio stations. The most recent Croatian song on Statehood Day 2008-2013 is analyzed in order to gain insight into their (memory) potential in terms of storing, interpreting and presenting a national identity. The research starts with the general assumption that popular music is an efficient identifier, transmitter, and promoter of national identity. The aim of the scientific research of the database was to analytically reveal specific titles of Croatian popular songs that participate in marking memories and analyzing their symbolic capital to gain insight into the popular music experience of the past and to develop a new method of scientifically based analysis of specific databases.

Keywords: specific databases, popular radio music, collective memory, national identity

Procedia PDF Downloads 328

Authors: Gregory W. Smith, Paul J. Riccomini

Abstract:

The goal of this empirical inquiry is to explore the effect of noise reducing headphones on the short-term memory recall of college students. Immediately following the presentation (via PowerPoint) of 12 unrelated and randomly selected one- and two-syllable words, students were asked to recall as many words as possible. Using a linear model with conditions marked with binary indicators, we examined the frequency and accuracy of words that were recalled. The findings indicate that for some students, a reduction of noise has a significant positive impact on their ability to recall information. As classrooms become more aurally distracting due to the implementation of cooperative learning activities, these findings highlight the need for a quiet learning environment for some learners.

Keywords: auditory distraction, education, instruction, noise, working memory

Procedia PDF Downloads 297

Authors: Firas Gerges, Frank Y. Shih

Abstract:

Malignant melanoma, known simply as melanoma, is a type of skin cancer that appears as a mole on the skin. It is critical to detect this cancer at an early stage because it can spread across the body and may lead to the patient's death. When detected early, melanoma is curable. In this paper, we propose a deep learning model (convolutional neural networks) in order to automatically classify skin lesion images as malignant or benign. Images underwent certain pre-processing steps to diminish the effect of the normal skin region on the model. The result of the proposed model showed a significant improvement over previous work, achieving an accuracy of 97%.

Keywords: deep learning, skin cancer, image processing, melanoma

Procedia PDF Downloads 111