Search

This paper proposes a real-time multi-scale semantic segmentation network (MsNet). MsNet is a combination of our novel multi-scale fusion with matching attention model (MFMA) as the decoding network and the network searched by asymptotic neural architecture search (ANAS) or MobileNetV3 as the encoding network. The MFMA not only extracts low level spatial features from multi-scale inputs but also decodes the contextual features extracted by ANAS. Specifically, considering the advantages and disadvantages of the addition fusion and concatenation fusion, we design multi-scale fusion (MF) that balances speed and accuracy. Then we creatively design two matching attention mechanisms (MA), including matching attention with low calculation (MALC) mechanism and matching attention with strong global context modeling (MASG) mechanism, to match varying resolutions and information of features at different levels of a network. Besides, the ANAS performs the deep neural network search by employing an asymptotic method and provide an efficient encoding network for MsNet, releasing researchers from those tedious mechanical trials. Through extensive experiments, we prove that MFMA, which can be applied to numerous recognition tasks, possesses excellent decoding ability. And we demonstrate the effectiveness and necessity of implementing the ``matching'' attention mechanism. Finally, the proposed two versions, MsNet₋ANAS and MsNet₋M achieve a new state-of-the-art trade-off between accuracy and speed on the CamVid and Cityscapes datasets. More remarkably, on the Nvidia Tesla V100 GPU, our MsNet₋ANAS achieves 74.1% mIoU with the speed of 184.2 FPS on the CamVid while 72.9% mIoU with the speed of 119.9 FPS on the Cityscapes.

Knowledge graph completion (KGC) can solve the problem of data sparsity in the knowledge graph. A large number of models for the KGC task have been proposed in recent years. However, the underutilisation of the structure information around nodes is one of the main problems of the previous KGC model, which leads to relatively single encoding information. To this end, a new KGC model that encodes and decodes the feature information is proposed. First, we adopt the subgraph sampling method to extract node structure. Moreover, the Graph convolutional network (GCN) introduced the channel attention convolution encode node structure features and represent them in matrix form to fully mine the node feature information. Eventually, the high-dimensional structure analysis weight decodes the encoded matrix embeddings and then constructs the scoring function. The experimental results show that the model performs well on the datasets used.

This letter designs a simultaneous lightwave and power transfer (SLIPT) system in the underwater environment using the power splitting technique. The data is sent by a laser diode and received by a multi-element receiver composed of a Single Photon Avalanche Diode (SPAD) and a solar cell that decodes the transmitted symbols and harvests electrical energy from the optical signal, respectively. In this context, we propose an energy-aware SLIPT system that optimizes the power split between the receivers and the SPAD detecting threshold to maximize the harvested energy while considering minimum link quality for the communication service. Although the optimization problem is not convex, we derive its feasibility conditions in terms of the system’s parameters and propose an optimal resource allocation algorithm. Finally, we introduce some numerical results to investigate the effect of SPAD’s dead time and other parameters on the system’s performance.

In this paper, we investigate the physical-layer security for an Internet of Things (IoT) relaying network employing non-orthogonal multiple access (NOMA) in the context of non-linear energy harvesting. In particular, a power-constrained source transmits its confidential information to a destination via an IoT device, which first decodes the received signal and then forwards the decoded data together with its own information using the NOMA technique to their respective destinations, while an eavesdropper is overhearing this transmission. Considering the scenario that both the power-constrained source and the IoT device are assumed to work under non-linear energy harvesting modes, and are capable of harvesting energy from a power beacon with a time switching protocol, both analytical and asymptotic expressions for secrecy outage probability as well as the analytical expression for the probability of strictly positive secrecy capacity are derived. In addition, those expressions are also verified with Monte-Carlo simulations.

Recovering information encodings from charts is to extract textual and graphical information from the diverse representations of charts to aid various applications that entail the hidden and directly available information. This paper proposes an end-to-end relationship-aware model for evaginating scientific charts for which an automated encoder-decoder framework is adopted. The framework integrates the CNN-LSTM model to extract direct and derived data by the influence of semantic relationships between the textual and graphical components. Semantic relationship makes the model robust towards diverse chart structures and orientations. Entity relationship-aware module extracts and builds relations amid textual-graphical components of the charts, and decodes the hidden data from the chart images. Our framework is one of its kind to recover data encoding from chart images based upon their inter-object semantic relationships, to the best of our knowledge. Model is tested upon public datasets, obtaining more than 97% accuracy compared with benchmark systems.

BackgroundTechnology to restore the ability to communicate in paralyzed persons who cannot speak has the potential to improve autonomy and quality of life. An approach that decodes words and sentences directly from the cerebral cortical activity of such patients may represent an advancement over existing methods for assisted communication.MethodsWe implanted a subdural, high-density, multielectrode array over the area of the sensorimotor cortex that controls speech in a person with anarthria (the loss of the ability to articulate speech) and spastic quadriparesis caused by a brain-stem stroke. Over the course of 48 sessions, we recorded 22 hours of cortical activity while the participant attempted to say individual words from a vocabulary set of 50 words. We used deep-learning algorithms to create computational models for the detection and classification of words from patterns in the recorded cortical activity. We applied these computational models, as well as a natural-language model that yielded next-word probabilities given the preceding words in a sequence, to decode full sentences as the participant attempted to say them.ResultsWe decoded sentences from the participant's cortical activity in real time at a median rate of 15.2 words per minute, with a median word error rate of 25.6%. In post hoc analyses, we detected 98% of the attempts by the participant to produce individual words, and we classified words with 47.1% accuracy using cortical signals that were stable throughout the 81-week study period.ConclusionsIn a person with anarthria and spastic quadriparesis caused by a brain-stem stroke, words and sentences were decoded directly from cortical activity during attempted speech with the use of deep-learning models and a natural-language model. (Funded by Facebook and others; ClinicalTrials.gov number, NCT03698149.).

This paper considers a cooperative device-to-device (D2D) system with non-orthogonal multiple access (NOMA). We assume that the base station (BS) can simultaneously communicate with all users to satisfy the full information transmission requirement. In order to characterize the impact of the weak channel and different decoding schemes, two novel decoding strategies are introduced: single signal decoding scheme and maximum ratio combining (MRC) decoding scheme , respectively. With the single signal decoding scheme , the users decode the received signals immediately after the receptions from the BS. On the other hand, the MRC decoding scheme jointly decodes the received signals via MRC until the corresponding phase comes and the users jointly decode the received signals by employing MRC. Considering Rayleigh fading channels, the ergodic sum-rate (SR), outage probability and outage capacity of the proposed D2D-NOMA system are analyzed. Moreover, approximate expressions for the ergodic SR are also provided with a negligible performance loss. Numerical results demonstrate that the ergodic SR and outage probability of the proposed D2D-NOMA scheme overwhelm that of the conventional NOMA schemes. Furthermore, it is also revealed that the system performance including the ergodic SR and outage probability are limited by the weak channel for both the single signal decoding scheme and conventional NOMA schemes, but not for the MRC decoding scheme .

In this paper, we suggest the use of Reconfigurable Intelligent Surfaces (RIS) for hybrid Radio Frequency (RF) Visible Light Communications (VLC). The signal is transmitted from a transmitter T to a relay C using RIS. RIS is implemented as a transmitter or a reflector between T and C. The relay node C decodes the received signal and transmits it to the receiver R using a VLC link. The proposed VLC/RF communications using RIS offers 20–45 dB gain with respect to conventional RF/VLC when the number of reflectors N is varied from $$N=16$$ to $$N=256$$ .

Brain-machine interfaces (BMIs) are useful tools for controlling assistive devices. One current approach in this field is continuous trajectory reconstruction (CTR), which decodes variables from electroencephalographic (EEG) signals. In this study, the CTR approach was applied to estimate kinetic variables from lower limb joints during pre-gait movements. The methodology consisted of a multi-layer perceptron capable of detecting patterns that represent the relation between EEG and torque signals acquired during motion tasks. The study also searched for an optimal subset of EEG channels based on the resulting patterns. A group of volunteers was recruited to execute pre-gait movements with both lower limbs to test the methodology. Results show that decodings of extracted patterns were successful in terms of the selected performance metrics; that is, the coefficient of determination, correlation coefficient, and signal-to-noise ratio. These metrics suggest that decodings for the right lower limb are better than for the left one, and that the most frequent electrodes in the optimal subsets for each task and joint show a tendency to lateralize. The results obtained suggest that this machine learning scheme could be used for future studies of the CTR of kinetic variables.

In this letter, we propose a media-based modulation (MBM)-aided uplink cooperative non-orthogonal multiple access (NOMA) scheme operating under hardware impairments and imperfect channel state information (CSI). In this scheme, the signal of the far user is transmitted with conventional modulation via the best pattern created from the ON/OFF status of the radio frequency (RF) mirrors. The near user decodes the far user's signal and transmits with the MBM technique by superposing its information. The bit error rate (BER) expressions of the proposed system for both users are derived under hardware impairments and imperfect CSI. Simulation results validate the derived BER expressions. Moreover, the proposed system is compared with systems in which the spatial modulation (SM) aided NOMA, generalized SM aided NOMA, and conventional NOMA applied at the near user. It has been shown that the proposed system is more robust to hardware impairments and imperfect CSI and outperforms these systems.

According to the atmospheric physical model, we can use accurate transmittance and atmospheric light information to convert a hazy image into a clean one. The scene-depth information is very important for image dehazing due to the transmittance directly corresponds to the scene depth. In this paper, we propose a multi-scale depth information fusion network based on the U-Net architecture. The model uses hazy images as inputs and extracts the depth information from these images; then, it encodes and decodes this information. In this process, hazy image features of different scales are skip-connected to the corresponding positions. Finally, the model outputs a clean image. The proposed method does not rely on atmospheric physical models, and it directly outputs clean images in an end-to-end manner. Through numerous experiments, we prove that the multi-scale deep information fusion network can effectively remove haze from images; it outperforms other methods in the synthetic dataset experiments and also performs well in the real-scene test set.

Bit-Patterned Magnetic Recording (BPMR) is one of the emerging data storage technologies, which promises an Areal Density (AD) of about 4 Tb/in2. However, a major problem practically encountered in a BPMR system is Inter-Track Interference (ITI) that can deteriorate the overall system performance, especially at high ADs. This paper proposes an iterative ITI cancellation method for an m-head m-track BPMR system, which uses m heads to read m adjacent tracks and decodes them simultaneously. To cancel the ITI, we subtract the weighted readback signals of adjacent tracks, acting as the ITI signals, from the readback signal of the target track, before passing the refined readback signal to a turbo decoder. Then, the decoded data will be employed to reconstruct the ITI signal for the next turbo iteration. Experimental results indicate that the proposed system performs better than the conventional system that uses one head to read one data track. Furthermore, we also find out that the proposed system is more robust to media noise and track misregistration than the conventional system.

Decoding the motion intention of the human upper limb from electroencephalography (EEG) signals has important practical values. However, existing decoding models are built under the attended state while subjects perform motion tasks. In practice, people are often distracted by other tasks or environmental factors, which may impair decoding performance. To address this problem, in this paper, we propose a hierarchical decoding model of human upper limb motion intention from EEG signals based on attention state estimation. The proposed decoding model includes two components. First, the attention state detection (ASD) component estimates the attention state during the upper limb movement. Next, the motion intention recognition (MIR) component decodes the motion intention by using the decoding models built under the attended and distracted states. The experimental results show that the proposed hierarchical decoding model performs well under the attended and distracted states. This work can advance the application of human movement intention decoding and provides new insights into the study of brain-machine interfaces.

Covert channels allow one to transmit information using mechanisms that were not originally intended for transmission. An example is a process in which a transmitter encodes information in moves of a character of a multiplayer game, and a receiver observes the moves and decodes the original message. This channel may be noisy, since the character may fall out of the receiver’s sight, a number of network packets may be lost, etc. Thus there emerges a natural problem of ogranizing a reliable channel. We propose a formal model called a partial erasure channel that describes the interaction of a transmitter and a receiver, introduce the notion of a consistent transmission protocol, formulate and prove the consistency criterion on the transmitting side and construct the optimal receiver for the given consistent transmitter.

This work describes the log-MAP (BCJR) algorithm implementation of a close to capacity dirty paper coding CODEC. The CODEC consists of eight deep pipeline processors. It decodes blocks of 975 bits in 26.9 ms using less than 9.7% of low-cost FPGA (and no DSP blocks). Two pipelines, for alpha and beta, calculate the values of gamma (of the BCJR) to reduce the storage requirements. The final log-likelihood ratio (LLR) is calculated together with alpha, reusing intermediate results. The number of bits used by the different signals of the processor is easily configurable. It was set to six bits to the channel measure signals and eight bits to log of probability signals like alpha, beta, and others. The CODEC clock was 100 MHz. The achieved bit rate is 36.2 Kbps per CODEC, but multiple CODECs can be fit into a single chip. The CODEC is 3.49 dB from the channel capacity.

The purpose of this project is to improve the control and precision of any rover by replacing the conventional DC motors or servo motors with the BLDC motors. The BLDC motors are much accurate to control because of the sensors they are equipped with. The telephonic signal from the receiver generates two frequencies one is a higher frequency and the other is a lower frequency. The higher frequency is used to drive the motors. The signal is then received by the receiver phone which then transmits the DTMF signal to the DTMF decoder. The DTMF decoder then decodes the signal and then transmits to the Gate Driver IC (which is connected to the Arduino for motor control). The Gate Driver IC which has the BLDC motors connected to it drives the motors. Furthermore this rover is being operated by Dual Tone Modulation Frequency (DTMF). In addition the rover is equipped with a Gas Sensor.

State monitoring of the complex system needs a large number of sensors. Especially, studies in soft electronics aim to attain complete measurement of the body, mapping various stimulations like temperature, electrophysiological signals, and mechanical strains. However, conventional approach requires many sensor networks that cover the entire curvilinear surfaces of the target area. We introduce a new measuring system, a novel electronic skin integrated with a deep neural network that captures dynamic motions from a distance without creating a sensor network. The device detects minute deformations from the unique laser-induced crack structures. A single skin sensor decodes the complex motion of five finger motions in real-time, and the rapid situation learning (RSL) ensures stable operation regardless of its position on the wrist. The sensor is also capable of extracting gait motions from pelvis. This technology is expected to provide a turning point in health-monitoring, motion tracking, and soft robotics., Real-time monitoring human motions normally demands connecting a large number of sensors in a complicated network. To make it simpler, Kim et al. decode the motion of fingers using a flexible sensor attached on wrist that measures skin deformation with the help of a deep-learning architecture.

Text generation tasks aim at generating human-readable text from different kinds of data. Normally, the generated text only contains the information included in the data and its application is thus restricted to some limited scenarios. In this paper, we extend the task to an open domain event text generation scenario with an entity chain as its skeleton. Specifically, given an entity chain containing several related event entities, the model should retrieve from a trustworthy repository (e.g. Wikipedia) the detailed information of these entities and generate a description text based on the retrieved sentences. We build a new dataset called WikiEvent1 that provides 34K pairs of entity chain and its corresponding description sentences. To solve the problem, we propose a wiki augmented generator framework that contains an encoder, a retriever, and a decoder. The encoder encodes the entity chain into a hidden space while the decoder decodes from the hidden space and generates description text. The retriever retrieves relevant text from a trustworthy repository which provides more information for generation. To alleviate the overfitting problem, we propose a novel random drop component that randomly deletes words from the retrieved sentences making our model more robust for handling long input sentences. We apply the proposed model on the WikiEvent dataset and compare it with a few baselines. The experimental results show that our carefully-designed architecture does help generate better event text, and extensive analysis further uncovers the characteristics of the proposed task.

In this work, we propose a new framework, called Document Image Transformer (DocTr), to address the issue of geometry and illumination distortion of the document images. Specifically, DocTr consists of a geometric unwarping transformer and an illumination correction transformer. By setting a set of learned query embedding, the geometric unwarping transformer captures the global context of the document image by self-attention mechanism and decodes the pixel-wise displacement solution to correct the geometric distortion. After geometric unwarping, our illumination correction transformer further removes the shading artifacts to improve the visual quality and OCR accuracy. Extensive evaluations are conducted on several datasets, and superior results are reported against the state-of-the-art methods. Remarkably, our DocTr achieves 20.02% Character Error Rate (CER), a 15% absolute improvement over the state-of-the-art methods. Moreover, it also shows high efficiency on running time and parameter count. The results will be available at https://github.com/fh2019ustc/DocTr for further comparison., This paper has been accepted by ACM Multimedia 2021

We demonstrate an end-to-end intelligent system of short-video generation for live-streaming, namely "VideoDiscovery'', which aims to automatically produce batches of high-value short-videos by discovering and organizing highlight content for commodity delivery. Traditionally, production of high-value short-videos for live-streaming is cost-expensive and time-consuming, which also demands experienced editing skills. To this end, we construct this system with three modules: 1)Semantic segment structuring first decodes live-streaming into a series of semantic candidates including commodity, Q&A, action, multi-modal, etc. 2)Hierarchical search engine performs automatically searches for semantically matching candidate shots from scripts. 3)Script-aware shot assembly is formulated combination problem over a graph of shots, considering temporal constraints and candidate idioms. Specifically, given an input live-streaming, the recommended video results illustrate diverse visual-semantic content, and follow script guidelines. Currently, our system has been launched online for Taobao stores, which enables to generate appealing videos in minutes for advertising and recommendation. The entry of our system is available at https://discovery.aliyun.com/index.

Contour detection plays an important role in both academic research and real-world applications. As the basic building block of many applications, its accuracy and efficiency highly influence the subsequent stages. In this work, we propose a novel lightweight system for contour detection that achieves state-of-the-art performance while keeps ultra-slim model size. The proposed method is built on an efficient encoder in a bottom-up/top-down fashion. Specially, we propose a novel decoder that compresses side features from an encoder and effectively decodes compact contextual information for high-accurate boundary localization. Besides, we propose a novel loss function that is able to assist a model to produce crisp object boundaries. We conduct extensive experiments to demonstrate the effectiveness of the proposed system on the widely adopted benchmarks BSDS500 and Multi-Cue. The results show that our system achieves the same best performance, yet only consumes 3.3% computational cost (16.45GFlops VS. 499.15GFlops) and 2.35% model size (1.94M VS. 82.43M) of the SOTA detector RCF-ResNet101. In the meantime, our method outperforms a large portion of the recent top edge detectors by a clear margin.

The formation of the electrical circuit of the compression device of the input unit, which produces a discrete cosine transform, and the electrical circuit of the output unit, which decodes the pulses from neurons into a binary code with space-time transformation and compression, have been performed. For this purpose, software modules were created for the automatic building of electrical circuits of the input and output units of the neuroprocessor and their subsequent numerical simulation in the specialized software “MDC-SPICE”, which performs calculations of large electrical circuits containing memristor-diode crossbars. The current-voltage characteristic of the memristor and the dependence of its electrical variables on time were plotted using the software module for visualizing the MDC-SPICE simulation results of the model with linear drift of oxygen vacancies.

Morse code is not only an elementary way of communication, but also a very convenient way for visually impaired people to communicate. Morse code alphabets are a combination of dots and dashes or dits and dahs and are based on international standards. In this paper, we will introduce a Morse Tool aka Morse pen, which is composed of five keys—dots, dashes, space, reset, and send, to help the visually impaired. In accordance with International Morse code, users will first make combinations of dots and dashes. The Morse Tool then decodes the input with the help of an on-board microcontroller into ASCII and transmits the same wirelessly via Bluetooth to a paired Smartphone device. “Unwired lite,” an android application used to display the text in a Smartphone. Now the text is in the Smartphone, and it can be used by many applications like a standard text editor, a messaging platform like WhatsApp or email.

In smart transportation systems, the infrastructures such as traffic lights are expected to transmit different safety messages according to the receptors (i.e., vehicle, pedestrian, platoon, etc.). The focus of this study is the infrastructure-to-platooning scenario where the lead-vehicle receives the signal from the traffic light (I2V), decodes, and re-transmits it again using its taillights to the following vehicle (V2V). Normally, the lead-vehicle also communicates with nearby vehicles to avoid the crashes and/or sharing the primary public safety information. Therefore, to enable the infrastructures and the vehicle to efficiently perform their multi-connections, it becomes necessary to know what the optimum power for the I2V and V2V links is. In this paper, we answer this question by realistic modeling of the hybrid I2V-V2V system. We first propose a new I2V path loss model considering the inherent characteristics of commercial traffic lights. For V2V link, we utilize a recent path loss model obtained with the same approach and with commercial taillights as the transmitters. Then, we obtain the SNR of both I2V and V2V links which are moreover used to derive the optimal power allocations that minimize the end-to-end BER of the I2V-V2V system. The effect of system parameters such as I2V and V2V distances, total transmit power, and system bandwidth on the BER is further investigated.

Communication with the outside world or the caretakers is one of the major challenges for the people with disabilities. In this work, the authors propose a realtime algorithm to detect morse code from the eye blinks in a series from a live camera. The proposed algorithm detects eye landmarks and estimates the level of eye-opening using eye aspect ratio (EAR). The low-cost software decodes morse code precisely based on the duration of the eyes closed or opened which in turn is translated to English Language. So, this system is intended to provide an alternative form of communication for people with disabilities and to convey confidential messages.

This paper presents the automatic recognition of COVID-19 from coughing. In particular, it describes our contribution to the DiCOVA challenge - Track 1, which addresses such cough sound analysis for COVID-19 detection. Pathologically, the effects of a COVID-19 infection on the respiratory system and on breathing patterns are known. We demonstrate the use of breathing patterns of the cough audio signal in identifying the COVID-19 status. Breathing patterns of the cough audio signal are derived using a model trained with the subset of the UCL Speech Breath Monitoring (UCL-SBM) database. This database provides speech recordings of the participants while their breathing values are captured by a respiratory belt. We use an encoder-decoder architecture. The encoder encodes the audio signal into breathing patterns and the decoder decodes the COVID-19 status for the corresponding breathing patterns using an attention mechanism. The encoder uses a pre-trained model which predicts breathing patterns from the speech signal, and transfers the learned patterns to cough audio signals. With this architecture, we achieve an AUC of 64:42% on the evaluation set of Track 1. Copyright © 2021 ISCA.

Scene text recognition has been widely investigated in computer vision. In the literature, the encoder-decoder based framework, which first encodes image into feature map and then decodes them into corresponding text sequences, have achieved great success. However, this solution fails in low-quality images, as the local visual features extracted from curved or blurred images are difficult to decode into corresponding text. To address this issue, we propose a new framework for Scene Text Recognition (STR), named Non-Local Attentional Scene Text Recognizer (NASTER). We use ResNet with Global Context Block (GC block) to extract global visual features. The global context information is then captured in parallel using the self-attention module and finally decoded by a multi-layer attention decoder with an intermediate supervision module. The proposed method achieves the state-of-the-art performances on seven benchmark datasets, demonstrating the effectiveness of our approach.

This paper proposes the analysis of the power allocation (PA) technique in visible light - non orthogonal multiple access (VL-NOMA) communication systems using uniform probability distribution function (UPDF) for improved system performance. A light-emitting diode (LED) transmits data to two mobile users (near- and far-users) in the indoor environment. The issues of better reception in a room and in distances to detect the signal prompted this study. NOMA decodes the signals for achievable rate based on the inequality agreement of the two mobile users while UPDF concentrates on the percentage at which the two mobile users are placed within the room. We have developed a new approach named UPDF-NOMA to solve the problem for PA in the VL-NOMA scenario. The UPDF-NOMA approach will decide the detection area for transmitted signal in percentage within the system. We observe that UPDF-NOMA performs 60% detection of the signal at 2m to 2.1m distance in the system. The simulation analysis presents the detected signals using the MATLAB software tool.

Contemporary soft keyboards possess limitations: the lack of physical feedback results in an increase of typos, and the interface of soft keyboards degrades the utility of the screen. To overcome these limitations, we propose an Invisible Mobile Keyboard (IMK), which lets users freely type on the desired area without any constraints. To facilitate a data-driven IMK decoding task, we have collected the most extensive text-entry dataset (approximately 2M pairs of typing positions and the corresponding characters). Additionally, we propose our baseline decoder along with a semantic typo correction mechanism based on self-attention, which decodes such unconstrained inputs with high accuracy (96.0%). Moreover, the user study reveals that the users could type faster and feel convenience and satisfaction to IMK with our decoder. Lastly, we make the source code and the dataset public to contribute to the research community., Comment: Accepted by IJCAI 2021

In the grant-free massive machine-type communication (mMTC) scenario, a key challenge is the joint device activity detection and data decoding. The sporadic nature of mMTC makes compressed sensing a promising solution to the activity detection problem. However, the typical two-phase coherent transmission scheme, which divides channel training and data decoding into two separate phases, suffers performance losses, especially when only a few bits of data are transmitted by each active device. This paper follows a newly proposed non-coherent transmission scheme in which the data bits are embedded in the pilot sequences and the BS simultaneously detects active devices and decodes the embedded data bits without explicit channel estimation. To exploit statistical channel information and the specific structure of the sparsity pattern introduced by the non-coherent transmission scheme, i.e., only one row in each section can be non-zero, we propose a receiving method based on the approximate message passing (AMP) algorithm with non-separable minimum mean-squared error denoisers specifically designed for the problem. The corresponding state evolution equations, which can be used to predict the section error rate (SER) performance, is obtained and simplified under certain assumptions. We also derive closed-form expressions of the SER performance based on the state evolution results. Finally, numerical simulations are given to validate the accuracy of the performance analysis and to show the superiority of the proposed receiving method over the conventional method based on AMP with separable denoisers in the literature.

Ambient backscatter communication has shown great potential in the development of future wireless networks. It enables a backscatter transmitter (BTx) to send information directly to an adjacent receiver by modulating over ambient radio frequency (RF) carriers. In this letter, we consider a cooperative ambient backscatter communication system where a multi-antenna cooperative receiver separately decodes signals from an RF source and a BTx. Upper bounds of the ergodic rates of both links are derived. The power scaling laws are accordingly characterized for both the primary cellular transmission and the cooperative backscatter. The impact of additional backscatter link is also quantitatively analyzed. Simulation results are provided to verify the derived results.

The decoding of transcription factor (TF) binding signals in genomic DNA is a fundamental problem. Here we present a prediction model called BindSpace that learns to embed DNA sequences and TF labels into the same space. By training on binding data from hundreds of TFs and embedding over 1 M DNA sequences, BindSpace achieves state-of-the-art multiclass binding prediction performance, in vitro and in vivo, and can distinguish between signals of closely related TFs.

In this paper, two energy harvesting policies deploying in cooperative non-orthogonal multiple access (NOMA) systems are considered. After period of wireless power transfer, the NOMA users including cell-edge and cell-center users simultaneously transmit the superposition coded symbols to the base station (BS). In the last time slot, the BS decodes to achieve its signal based on superposition coded symbol with corresponding power allocation factors. This paper provides exact expressions of outage probability in two schemes. Performance gap of two NOMA users can be raised by providing different power allocation factors. It is confirmed by numerical result. Distance and data rate are main factors affecting outage performance. Scheme I exhibit scenario where power beacon transmits energy signal to NOMA user while the BS feeds energy to NOMA user in Scheme II. It is shown that outage performance of Scheme I is better than that of Scheme II. Web of Science 25 4 80 75

Wavelet forced de-noising algorithm is suitable for denoising of unsteady drilling fluid pulse signal, including baseline drift rectification and two-stage de-noising processing of frame synchronization signal and instruction signal. Two-stage de-noising processing can reduce the impact of baseline drift and determine automatic peak detection threshold range for signal recognition by distinguishing different features of frame synchronization pulse and instruction pulse. Rising and falling edge relative protruding threshold is defined for peak detection in signal recognition, which can make full use of the degree of the signal peak change and detect peaks flexibly with rising and falling edge relative protruding threshold combination. A synchronous decoding method was designed to reduce position uncertainty of the frame synchronization pulse and eliminate the accumulative error of time base drift, which determines the first instruction pulse position according to position of the frame synchronization pulse and decodes subsequent instruction pulse by taking current instruction pulse as new bit synchronization pulse. Special tool software was developed to tune algorithm parameters, which has a decoding success rate of about 95% for the universal coded signals. For the special coded signals with check byte, decoding success rate using the automatic threshold adjustment algorithm is as high as 99%. Key Words: drilling fluid, pulse signal, signal processing, decoding success rate, automatic de-noising and recognition, wavelet forced de-noising, peak detection, synchronous decoding

Molecular Communication via Diffusion (MCvD) is a promising paradigm which enables nano-machines to communicate with each other. However, the reliability of existing communication systems degrades rapidly as the distance between the transmitters and the receivers grows. To solve this issue, relaying schemes must be implemented in practice. In this paper, we study two relaying schemes: In the first case, the relay node decodes the incoming signal symbol and forwards it to the receiver using a different types of molecules. Then, the receiver detects the information bits by only considering the molecules from the relay node. In the second case, the receiver considers both types of molecules sent from the transmitter and the relay node. Closed-form expression for the end-to-end bit error probability of the these two schemes are derived. Also, we obtain optimal relay location for both schemes. Moreover, we study the optimal threshold, concentration, baud-rate and timing offset to mitigate the overall Inter-Symbol Interference(ISI) and obtain maximum throughput in the relay network.

The Viterbi algorithm is an efficient method to decode convolution encoded data using the concept of maximum likelihood estimation. This paper presents the design and FPGA implementation of a Viterbi decoder for satellite communication. The design is coded in Verilog HDL using the Vivado 2017.4 tool and QuestaSim for behavioral and post-layout simulations. The decoder is designed to decode the output of rate half convolution encoder. It successfully decodes data for any constraint length and corrects error up to 4 bits. The implementation is done on the Zynq-7000 development board. The maximum operating frequency achieved is 221.9 MHz with a power consumption of 37.62 mW.

Automatic building extraction from satellite RGB images, is a low-cost alternative to perform important urban planning tasks. Yet, it is a challenging one, especially when natural and non-city block objects interfere in the semantic segmentation of algorithms that extract their key features. In this work we approach the automatic building extraction using a Convolution Neural Network based on the U-Net architecture. In contrast to existing approaches, it successfully encodes important features and decodes the buildings’ localization by requiring both reduced computational time and dataset size. We evaluate the U-Net’s performance using RGB images selected from the SpaceNet 1 dataset and the experimental results show an accuracy in building localization of 92.3%. Finally, favorable comparison with existing CNN approaches to hyperspectral images targeting the SpaceNet 1 dataset, demonstrated its effectiveness.

Molecular communication is defined as the transmission and reception of biochemical information through molecules. Chemical signals are used to transfer information in molecular communication. In addition, carrier molecules are used to transmit and receive information between micro and nano machines. Molecular communication systems can be used in many subject such as Cancer Treatment, Drug Delivery. Molecular communication systems generally consist of a transmitter (Tx) and a receiver (Rx). Just like in digital communication, the transmitter transmits information through molecules, while the receiver receives and decodes the information encoded in the transmitter. In this study, a molecular communication model using mobile transmitters and transmitters is designed and analyzed in Matlab environment. In the designed model, the receptors on the receiver are considered as antennas and all calculations have been made accordingly. Contrary to the literature, comments have been made about how the communication quality will change with the change of distance between receiver and transmitter and diffusion constant of the nano machines in a mobile system. As a result, it has been observed that as the distance between the transmitter and receiver increases, the communication quality decreases.

Volitional control of prostheses is most commonly achieved by myoelectric signalling. The electromyograph (EMG) is detected and processed by a controller, that decodes and relates the signal to the corresponding position of the prosthetic. Myoelectric signalling is limited in users by two factors: lack of nerve endings corresponding to the position of the amputation, and neurological damage resulting in poor signal control. Improved prosthesis control has been demonstrated by the addition of feedback sensors based on computer vision and inertial measurement units. Computer vision requires a significant level of processing, resulting in a high latency and high power usage. In this paper, we propose a means of overcoming this limitation by use of in-vivo retinal signalling to complement EMG for improved control. This is demonstrated using a real-time conductance- based simulator as the sole method of control for an upper-limb prosthesis. Input image streams are received by a camera and used to activate the combined rod and cone photoreceptor cell responses. This in turn generates a spike train which is counted and averaged over time, and passed to an Arduino-based control system which modulates the behavior of the prosthesis. We seek to use this system to lower the experimental barriers of in-vivo ganglion electrical signalling by presenting a way to use retina emulation. A link to the simulator is provided.

Massive MIMO (mMIMO) communication has been considered as pivotal and significant technology to furnish the expected demand of Fifth-generation (5G) and beyond 5G systems. It provides enormous potentiality to 5G systems by the deployment of excessive antennas at base station. Therefore, the probability of using different modulation by different transmitters has been increased aiming to reduce bit error rate as well as enhance data rate. It leads to develop intelligent receiver that can efficiently classify and recognize accurate modulation and decodes data. To shed light on this issue, automatic modulation recognition scheme based on ensemble classifier has been studied in this paper. The overall performance analysis of the ensemble classifier for several modulated radio signal has been presented in this article. The simulation results justify the exactitude of ensemble classifier as efficient one for modulation classification.

Despite the promising features of neural interfaces, their trade-off between information transfer and invasiveness has limited translation and viability outside research settings. Here, we present a non-invasive neural interface that provides access to spinal motoneuron activities from a sensor band at the wrist. The interface decodes electric signals present at the tendon endings of the forearm muscles by using a model of signal generation and deconvolution. First, we evaluated the reliability of the interface to detect motoneuron firings, and thereafter we used the decoded neural activity for the prediction of finger movements in offline and real-time conditions. The results showed that motoneuron activity decoded from the wrist accurately predicted individual and combined finger commands and therefore allowed for highly accurate real-time control. These findings demonstrate the feasibility of a wearable, non-invasive, neural interface at the wrist for precise real-time control based on the output of the spinal cord.

In this paper, we build a multi-style generative model for stylish image captioning which uses multi-modality image features, ResNeXt features and text features generated by DenseCap. We propose the 3M model, a Multi-UPDOWN caption model that encodes multi-modality features and decode them to captions. We demonstrate the effectiveness of our model on generating human-like captions by examining its performance on two datasets, the PERSONALITY-CAPTIONS dataset and the FlickrStyle10K dataset. We compare against a variety of state-of-the-art baselines on various automatic NLP metrics such as BLEU, ROUGE-L, CIDEr, SPICE, etc. A qualitative study has also been done to verify our 3M model can be used for generating different stylized captions., To be published at FLAIRS-34

Direct decoding for task-oriented dialogue is known to suffer from the explaining-away effect, manifested in models that prefer short and generic responses. Here we argue for the use of Bayes' theorem to factorize the dialogue task into two models, the distribution of the context given the response, and the prior for the response itself. This approach, an instantiation of the noisy channel model, both mitigates the explaining-away effect and allows the principled incorporation of large pretrained models for the response prior. We present extensive experiments showing that a noisy channel model decodes better responses compared to direct decoding and that a two stage pretraining strategy, employing both open-domain and task-oriented dialogue data, improves over randomly initialized models., Accepted to TACL, pre MIT Press publication version

The purpose of this project is to improve the control and precision of any rover by replacing the convention direct current motors or servo motors with brushless direct current motors. The BLDC motors are much accurate to control because of the sensors they are equipped with. The telephonic signal from the transmitter generates two frequencies one is a lower frequency and the other is a higher frequency. The higher frequency is used to drive the motors. The signal is then sent to the receiver which then transmits the dual tone modulation frequency signal to the DTMF decoder. The DTMF decoder then decodes the received signal and then transmits to the gate driver IC (which is connected to the Arduino for motor control). The Gate Driver IC which has the BLDC motors connected to it drives the motors. Furthermore this rover is being operated by Dual Tone Modulation Frequency (DTMF). In addition the rover is equipped with a Gas Sensor.

In classical computers, information is represented as the binary digits 0 or 1. These are called bits. For example, the number 1 in an 8-bit binary representation is written as 00000001. The number 2 is represented as 00000010. We place extra zeros in front to write every number with 8-bits total, which is called one byte. In fact, every classical computer translates these bits into the human readable information on your electronic device. The document you read or video you watch is encoded in the computer binary language in terms of these 1’s and 0’s. Computer hardware understands the 1-bit as an electrical current flowing through a wire (in a transistor) while the 0-bit is the absence of an electrical current in a wire. These electrical signals can be thought of as “on” (the 1-bit) or “off” (the 0-bit). Your computer then decodes the classical 1 or 0 bits into words or videos, etc.

This chapters presents two ways to write tests for code that directly decodes JSON input and dives into the trade-offs of using TDD for implenting this kind of logic.

Power decoding is a partial decoding paradigm for arbitrary algebraic geometry codes for decoding beyond half the minimum distance, which usually returns the unique closest codeword, but in rare cases fails to return anything. The original version decodes roughly up to the Sudan radius, while an improved version decodes up to the Johnson radius, but has so far been described only for Reed--Solomon and one-point Hermitian codes. In this paper we show how the improved version can be applied to any algebraic geometry code.

Offline handwritten text recognition is still a big challenging problem due to various backgrounds, noises, diversity of writing styles, and multiple touches between characters. In this paper, we propose a model of 2D Self-Attention Convolutional Recurrent Network (2D-SACRN) for recognizing handwritten text lines. The 2D-SACRN model consists of three main components: 1) a 2D self-attention based convolutional feature extractor that extracts a feature sequence from an input image; 2) a recurrent encoder that encodes the feature sequence into a sequence of label probabilities; and 3) a CTC-decoder that decodes the sequence of label probabilities into the final label sequence. In this model, we present a 2D self-attention mechanism in the feature extractor to capture the relationships between widely separated spatial regions in an input image. In the experiment, we evaluate the performance of the proposed model on the three datasets: IAM Handwriting, Rimes, and TUAT Kondate. The experimental results show that the proposed model achieves similar or better accuracy when compared to state-of-the-art models in all datasets.