Decoding Phase Research Articles

Multiuser Detection for Downlink Communication in LoRa- Like Networks

Internet of Things (IoT) technology has become ubiquitous in a multitude of applications and its use is growing. However, the expansion of IoT faces a major difficulty: scalability, that is, very dense deployment of communicating devices is currently limited. In long-range networks, such as LoRa, the downlink is critical because it limits the number of acknowledgements that can be sent, and consequently reliability. It also limits the possibility to update the devices, which could be critical when they are deployed for decades. To overcome those problems, we propose a solution, inspired by Non Orthogonal Multiple Access (NOMA) techniques, to increase by at least one order of magnitude the number of devices that can be addressed. While the approach differentiates the devices by the power allocated to them, it differs from the vast majority of previous works on power domain NOMA because it does not require interference cancellation. Instead, it benefits from the spectrum spreading of the modulation scheme (chirp spread spectrum), where, at the end of the decoding phase, the information carried by a symbol is found in the position of a peak in the Fourier domain. In the vast majority of cases, the information from different users results in different peak positions, not creating any interference. In that sense, we get closer to avoidance schemes such as time or frequency hopping, but without using a code. In this article, we propose a new solution for NOMA in the power domain that does not suffer from the limitations induced by interference cancellation residues. The proposed system, including preamble detection and channel estimation, is presented and evaluated by simulations. We demonstrate that our proposed scheme increases the number of devices by one order of magnitude compared to the current system which allows addressing only one user at a time and maintains full compatibility with the LoRa physical layer standard.

Employing Differentiable Neural Computers for Image Captioning and Neural Machine Translation

Abstract In the history of artificial neural networks, LSTMs have proved to be a high-performance architecture at sequential data learning. Although LSTMs are remarkable in learning sequential data but are limited in their ability to learn long-term dependencies and representation of certain data structures because of the lack of external memory. In this paper, we tackled two main tasks, one is language translation and other is image captioning. We approached the problem of language translation by leveraging the capabilities of the recently developed DNC architectures. Here we modified the DNC architecture by including dual neural controllers instead of one and an external memory module. Inside our controller, we employed a neural network with memory-augmentation which differs from the original differentiable neural computer, we implemented a dual controller’s system in which one controller is for encoding the query sequence whereas another controller is for decoding the translated sequences. During the encoding cycle, new inputs are read and the memory is updated accordingly. In the decoding cycle, the memory is protected from any writing from the decoding controller. Thus, the decoder phase generates a translated sequence at a time step. Therefore, the proposed dual controller neural network with memory-augmentation is then trained and tested on the Europarl dataset. For the image captioning task, our architecture is inspired by an end-to-end image captioning model where CNN’s output is passed to RNN as input only once and the RNN generates words depending on the input. We trained our DNC captioning model on 2015 MSCOCO dataset. In the end, we compared and shows the superiority of our architecture as compared to conventionally used LSTM and NTM architectures.

Future-Aware Knowledge Distillation for Neural Machine Translation

Although future context is widely regarded useful for word prediction in machine translation, it is quite difficult in practice to incorporate it into neural machine translation. In this paper, we propose a future-aware knowledge distillation framework (FKD) to address this issue. In the FKD framework, we learn to distill future knowledge from a backward neural language model (teacher) to future-aware vectors (student) during the training phase. The future-aware vector for each word position is computed in a bridge network and optimized towards the corresponding hidden state in the backward neural language model via a knowledge distillation mechanism. We further propose an algorithm to jointly train the neural machine translation model, neural language model and knowledge distillation module end-to-end. The learned future-aware vectors are incorporated into the attention layer of the decoder to provide full-range context information during the decoding phase. Experiments on the NIST Chinese-English and WMT English-German translation tasks show that the proposed method significantly improves translation quality and word alignment.

Design and Implementation of a Polar Codes Blind Detection Scheme

In blind detection, a set of candidates has to be decoded within a strict time constraint, to identify which transmissions are directed at the user equipment. Blind detection is required by the 3GPP LTE/LTE-A and fifth generation (5G) standards. With the selection of polar codes in 5G, the issue of blind detection of polar codes needs to be addressed. A polar code blind detection scheme has been recently proposed where the user ID is transmitted instead of some of the frozen bits. We propose an architecture to implement an improved version of such scheme. A first, coarse decoding phase helps selecting a subset of candidates that is decoded by a more powerful algorithm: an early stopping criterion is also introduced for the second decoding phase. The architecture relies on a tunable decoder that can be used for both phases. The architecture is synthesized and implementation results are reported for various system parameters. The reported area occupation and latency, obtained in 65-nm CMOS technology, are able to meet 5G requirements.

Syntax-Based Chinese-Vietnamese Tree-to-Tree Statistical Machine Translation with Bilingual Features

Because of the scarcity of bilingual corpora, current Chinese--Vietnamese machine translation is far from satisfactory. Considering the differences between Chinese and Vietnamese, we investigate whether linguistic differences can be used to supervise machine translation and propose a method of syntax-based Chinese--Vietnamese tree-to-tree statistical machine translation with bilingual features. Analyzing the syntax differences between Chinese and Vietnamese, we define some linguistic difference-based rules, such as attributive position, time adverbial position, and locative adverbial position, and create rewards for similar rules. These rewards are integrated into the extraction of tree-to-tree translation rules, and we optimize the pruning of the search space during the decoding phase. The experiments on Chinese--Vietnamese bilingual sentence translation show that the proposed method performs better than several compared methods. Further, the results show that syntactic difference features, with search pruning, can improve the accuracy of machine translation without degrading the efficiency.

An Annotation Model on End-to-End Chest Radiology Reports

Annotating radiographic images with tags is an indispensable preliminary work in computer-aided medical research, which requires professional physician participated in and is quite time-consuming. Therefore, how to automatically annotate radiographic images has become the focus of researchers. However, image report texts, containing crucial radiologic information, have not to be given enough attention for images annotation. In this paper, we propose a neural sequence-to-sequence annotation model. Especially, in the decoding phase, a probability is first learned to copy existing words from report texts or generate new words. Second, to incorporate the patient’s background information, “indication” section of the report is encoded as a sentence embedding, and concatenated with the decoder neural unit input. What’s more, we devise a more reasonable evaluation metric for this annotation task, aiming at assessing the importance of different words. On the Open-i dataset, our model outperforms existing non-neural and neural baselines under the BLEU-4 metrics. To our best knowledge, we are the first to use sequence-to-sequence model for radiographic image annotation.

NADAQ: Natural Language Database Querying Based on Deep Learning

The high complexity behind SQL language and database schemas has made database querying a challenging task to human programmers. In this paper, we present our new natural language database querying (NADAQ) system as an alternative solution, by designing new translation models smoothly fusing deep learning and traditional database parsing techniques. On top of the popular encoder-decoder model for machine translation, NADAQ injects new dimensions of schema-aware bits associated with the input words into encoder phase and adds new hidden memory neurons controlled by the finite state machine for grammatical state tracking into the decoder phase. We further develop new techniques to enable the augmented neural network to reject queries irrelevant to the contents of the target database and recommend candidate queries reversely transformed into natural language. NADAQ performs well on real-world database systems over human labeled workload, returning query results at 90% accuracy.

Modeling Past and Future for Neural Machine Translation

Existing neural machine translation systems do not explicitly model what has been translated and what has not during the decoding phase. To address this problem, we propose a novel mechanism that separates the source information into two parts: translated Past contents and untranslated Future contents, which are modeled by two additional recurrent layers. The Past and Future contents are fed to both the attention model and the decoder states, which provides Neural Machine Translation (NMT) systems with the knowledge of translated and untranslated contents. Experimental results show that the proposed approach significantly improves the performance in Chinese-English, German-English, and English-German translation tasks. Specifically, the proposed model outperforms the conventional coverage model in terms of both the translation quality and the alignment error rate.

Assessing the Relevance of Specific Response Features in the Neural Code.

The study of the neural code aims at deciphering how the nervous system maps external stimuli into neural activity—the encoding phase—and subsequently transforms such activity into adequate responses to the original stimuli—the decoding phase. Several information-theoretical methods have been proposed to assess the relevance of individual response features, as for example, the spike count of a given neuron, or the amount of correlation in the activity of two cells. These methods work under the premise that the relevance of a feature is reflected in the information loss that is induced by eliminating the feature from the response. The alternative methods differ in the procedure by which the tested feature is removed, and the algorithm with which the lost information is calculated. Here we compare these methods, and show that more often than not, each method assigns a different relevance to the tested feature. We demonstrate that the differences are both quantitative and qualitative, and connect them with the method employed to remove the tested feature, as well as the procedure to calculate the lost information. By studying a collection of carefully designed examples, and working on analytic derivations, we identify the conditions under which the relevance of features diagnosed by different methods can be ranked, or sometimes even equated. The condition for equality involves both the amount and the type of information contributed by the tested feature. We conclude that the quest for relevant response features is more delicate than previously thought, and may yield to multiple answers depending on methodological subtleties.

Chinese remainder theorem-based two-in-one image secret sharing with three decoding options

Two-in-one image secret sharing (ISS) with more decoding options has more applicable scenarios than that with only one decoding option. Conventional two-in-one ISS methods with more decoding options mainly combine visual secret sharing (VSS) and polynomial-based or XOR-based ISS. They have the drawbacks of lossy recovery for (k,n) threshold, only two decoding options, high computation in recovering phase or large pixel expansion. In this article, we design a Chinese remainder theorem (CRT)-based two-in-one ISS with three decoding options: lossless recovery, grayscale stacking recovery and visual previewing capability. In the encoding phase, we encode a binary secret image by random grid (RG)-based VSS (RGVSS). Meanwhile, a grayscale secret image is encoded by CRT in the processing of encoding the binary secret image. In the decoding phase, we can decode the binary secret image with two decoding options: grayscale stacking recovery and visual previewing capability. The grayscale secret image can be losslessly decoded with one decoding option through solving a set of linear congruence equations with only modular operation. Furthermore, the shadow image has no pixel expansion. We show our effectiveness by means of theoretical analyses and experiments.

Table-to-Text Generation by Structure-Aware Seq2seq Learning

Table-to-text generation aims to generate a description for a factual table which can be viewed as a set of field-value records. To encode both the content and the structure of a table, we propose a novel structure-aware seq2seq architecture which consists of field-gating encoder and description generator with dual attention. In the encoding phase, we update the cell memory of the LSTM unit by a field gate and its corresponding field value in order to incorporate field information into table representation. In the decoding phase, dual attention mechanism which contains word level attention and field level attention is proposed to model the semantic relevance between the generated description and the table. We conduct experiments on the WIKIBIO dataset which contains over 700k biographies and corresponding infoboxes from Wikipedia. The attention visualizations and case studies show that our model is capable of generating coherent and informative descriptions based on the comprehensive understanding of both the content and the structure of a table. Automatic evaluations also show our model outperforms the baselines by a great margin. Code for this work is available on https://github.com/tyliupku/wiki2bio.

Translation in Prokaryotes.

This review summarizes our current understanding of translation in prokaryotes, focusing on the mechanistic and structural aspects of each phase of translation: initiation, elongation, termination, and ribosome recycling. The assembly of the initiation complex provides multiple checkpoints for messenger RNA (mRNA) and start-site selection. Correct codon-anticodon interaction during the decoding phase of elongation results in major conformational changes of the small ribosomal subunit and shapes the reaction pathway of guanosine triphosphate (GTP) hydrolysis. The ribosome orchestrates proton transfer during peptide bond formation, but requires the help of elongation factor P (EF-P) when two or more consecutive Pro residues are to be incorporated. Understanding the choreography of transfer RNA (tRNA) and mRNA movements during translocation helps to place the available structures of translocation intermediates onto the time axis of the reaction pathway. The nascent protein begins to fold cotranslationally, in the constrained space of the polypeptide exit tunnel of the ribosome. When a stop codon is reached at the end of the coding sequence, the ribosome, assisted by termination factors, hydrolyzes the ester bond of the peptidyl-tRNA, thereby releasing the nascent protein. Following termination, the ribosome is dissociated into subunits and recycled into another round of initiation. At each step of translation, the ribosome undergoes dynamic fluctuations between different conformation states. The aim of this article is to show the link between ribosome structure, dynamics, and function.

A Review of Image Compression Using Fractal Image Compression with Neural Network

Generally the fractal image compression is a new process in the images compression. It is a block based image compression technique, which detects and decodes the existing similarities between different regions in the image. The main disadvantage of FIC is that the encoding time is comparatively very high, w here as the decoding time is very short. An artificial intelligence technique like neural network is used to reduce the search space and encoding time for the MRI images with an algorithm called as “back propagation” neural network algorithm. Initially, MRI image is divided into ranges and domains of fixed size. The best matched domain is selected for each range block and its range index and best matched domain index are produced, which acts as input to the expert system and which results reduced the sets of matched domain blocks. The neural network is then trained with these resultant values. This trained net is now used to compress other MRI images which lead to a very less encoding time. During the decoding phase, the transformation parameters are recursively applied to any random original image, which then converges to the fractal image after some changes. The simulation results show that the performance of this Neural Network based FIC is really. This paper shows the neural network based FIC which produces high development in encoding time without corrupting the image quality when compared to normal FIC.

A Review of Image Compression Using Fractal Image Compression with Neural Network

Generally the fractal image compression is a new process in the images compression. It is a block based image compression technique, which detects and decodes the existing similarities between different regions in the image. The main disadvantage of FIC is that the encoding time is comparatively very high, w here as the decoding time is very short. An artificial intelligence technique like neural network is used to reduce the search space and encoding time for the MRI images with an algorithm called as “back propagation” neural network algorithm. Initially, MRI image is divided into ranges and domains of fixed size. The best matched domain is selected for each range block and its range index and best matched domain index are produced, which acts as input to the expert system and which results reduced the sets of matched domain blocks. The neural network is then trained with these resultant values. This trained net is now used to compress other MRI images which lead to a very less encoding time. During the decoding phase, the transformation parameters are recursively applied to any random original image, which then converges to the fractal image after some changes. The simulation results show that the performance of this Neural Network based FIC is really. This paper shows the neural network based FIC which produces high development in encoding time without corrupting the image quality when compared to normal FIC.

Blind Detection With Polar Codes

In blind detection, a set of candidates has to be decoded within a strict time constraint, to identify which transmissions are directed at the user equipment. We propose a blind detection scheme based on polar codes, where the radio network temporary identifier is transmitted instead of some of the frozen bits. A low-complexity decoding phase decodes all candidates, selecting a subset that is decoded by a high-performance algorithm. Simulations results show good missed detection and false alarm rates, that meet the 3GPP LTE-A and future 5G standard specifications. We also propose an early stopping criterion that can reduce the number of operations performed, improving both average latency and energy consumption. The detection speed is analyzed and different system parameter combinations are shown to meet the stringent timing requirements, leading to various implementation trade-offs.

Separable reversible data hiding and encryption for HEVC video

A separable reversible data hiding and encryption for high efficiency video coding (HEVC) video is proposed in this paper. In the encoding phase, one can encrypt the signs and amplitudes of motion vector differences and the signs of residual coefficients with an encryption key by using RC4, while another can hide data into nonzero AC residual coefficients with a hiding key. For the decoding phase, one can decrypt the HEVC video and obtain a HEVC video similar to the original one if he only has the encryption key; he can extract the hiding data, but he does not know the HEVC video content if he only has the hiding key; he can extract the hiding data and recover the original HEVC video if he has both the encryption key and hiding key. Experimental results and analysis show that the proposed scheme can obtain a good security performance and keep the HEVC video stream format compliant.

Securing Coding-Based Cloud Storage Against Pollution Attacks

The widespread diffusion of distributed and cloud storage solutions has changed dramatically the way users, system designers, and service providers manage their data. Outsourcing data on remote storage provides indeed many advantages in terms of both capital and operational costs. The security of data outsourced to the cloud, however, still represents one of the major concerns for all stakeholders. Pollution attacks, whereby a set of malicious entities attempt to corrupt stored data, are one of the many risks that affect cloud data security. In this paper we deal with pollution attacks in coding-based block-level cloud storage systems, i.e., systems that use linear codes to fragment, encode, and disperse virtual disk sectors across a set of storage nodes to achieve desired levels of redundancy, and to improve reliability and availability without sacrificing performance. Unfortunately, the effects of a pollution attack on linear coding can be disastrous, since a single polluted fragment can propagate pervasively in the decoding phase, thus hampering the whole sector. In this work we show that, using rateless codes, we can design an early pollution detection algorithm able to spot the presence of an attack while fetching the data from cloud storage during the normal disk reading operations. The alarm triggers a procedure that locates the polluting nodes using the proposed detection mechanism along with statistical inference. The performance of the proposed solution is analyzed under several aspects using both analytical modelling and accurate simulation using real disk traces. Our results show that the proposed approach is very robust and is able to effectively isolate the polluters, even in harsh conditions, provided that enough data redundancy is used.

Recovery of Watermarked Image from Geometrics Attacks using Effective Histogram Shape based Index

Objective: The main objective of the paper is to provide security and proof of ownership to the watermarked image. Methods: Watermark is a transparent or an undetectable signature implanted inside an image to show uniqueness or proof of ownership. The illegally used images can undergo different kind of attacks namely signal processing attacks and geometric attacks. We have used Tree Based Parity Check (TBPC) algorithm to deal with these common attacks. We preprocess the host image using Gaussian filter. And using Fast Walsh-Hadamard method, we embed watermark inside an Image. Findings: We choose the Gray levels of the image by the means of secret key. Later we construct the histogram based on the gray level that is chosen. In these chosen pixels group’s watermarks are embedded. Then a secret text or code is inserted into this filtered image. To encrypt this text we use Tree Based Parity Check algorithm. Based on the secret key that is available, in the decoding phase we identify watermark pixel group and then we extract the watermarks from them. Application: The proposed system can achieve strong robustness of the image and detailed view of Bone structure in X-ray images. It can also be used to find the duplicate currency notes.

GST

Modern GPUs supporting compressed textures allow interactive application developers to save scarce GPU resources such as VRAM and bandwidth. Compressed textures use fixed compression ratios whose lossy representations are significantly poorer quality than traditional image compression formats such as JPEG. We present a new method in the class of supercompressed textures that provides an additional layer of compression to already compressed textures. Our texture representation is designed for endpoint compressed formats such as DXT and PVRTC and decoding on commodity GPUs. We apply our algorithm to commonly used formats by separating their representation into two parts that are processed independently and then entropy encoded. Our method preserves the CPU-GPU bandwidth during the decoding phase and exploits the parallelism of GPUs to provide up to 3X faster decode compared to prior texture supercompression algorithms. Along with the gains in decoding speed, our method maintains both the compression size and quality of current state of the art texture representations.

Asymmetrical modulation of time perception by increase versus decrease in coherence of motion.

Stimulus properties are known to affect duration judgments. In this study, we tested the effect of motion coherence levels in randomly moving dots on the perceived duration of these stimuli. In Experiments 1 and 2 we tested participants on a temporal reproduction task, using stimuli with varying degrees of motion coherence as the to-be-timed stimuli. Our results in both experiments showed that increasing motion coherence from the encoded (i.e. the first) to the reproduced (i.e. the second) stimulus leads to longer reproduction times. These effects were primarily additive in nature, and their magnitude increased with the difference between the coherence levels in the encoding versus reproduction (decoding) phases. This effect was not mirrored when there was a decrease in motion coherence. Experiment 3 tested if the differential number of exploratory saccadic eye-movements during encoding and reproduction predicted these effects. The behavioral findings of Experiment 1 and 2 were replicated in the third experiment, and the change in the number of eye movements from encoding to reproduction predicted the reproduction time when there was an increase in motion coherence. These results are explained by the effect of attention on the latency to initiate temporal integration that is only manifested when there is an increase in the level of motion coherence.