Practically, it is more feasible to collect compact visual features rather than the video streams from hundreds of thousands of cameras into the cloud for big data analysis and retrieval. Then the problem becomes whic...
详细信息
Practically, it is more feasible to collect compact visual features rather than the video streams from hundreds of thousands of cameras into the cloud for big data analysis and retrieval. Then the problem becomes which kinds of features should be extracted, compressed and transmitted so as to meet the requirements of various visual tasks. Recently, many studies have indicated that the activations from the convolutional layers in convolutional neural networks (CNNs) can be treated as local deep features describing particular details inside an image region, which are then aggregated (e.g., using Fisher Vectors) as a powerful global descriptor. Combination of local and global features can satisfy those various needs effectively. It has also been validated that, if only local deep features are coded and transmitted to the cloud while the global features are recovered using the decoded local features, the aggregated global features should be lossy and consequently would degrade the overall performance. Therefore, this paper proposes a joint coding framework for local and global deep features (DFJC) extracted from videos. In this framework, we introduce a coding scheme for real-valued local and global deep features with intra-frame lossy coding and inter-frame reference coding. The theoretical analysis is performed to understand how the number of inliers varies with the number of local features. Moreover, the inter-feature correlations are exploited in our framework. That is, local feature coding can be accelerated by making use of the frame types determined with global features, while the lossy global features aggregated with the decoded local features can be used as a reference for global feature coding. Extensive experimental results under three metrics show that our DFJC framework can significantly reduce the bitrate of local and global deep features from videos while maintaining the retrieval performance.
Information metasurfaces have attracted much attention in recent years due to the capability to link the physical world and information science. However, most of the current information metasurfaces are either phase-o...
详细信息
Information metasurfaces have attracted much attention in recent years due to the capability to link the physical world and information science. However, most of the current information metasurfaces are either phase-only coding or amplitude-only coding, limiting their functions and applications. Here, a broadband and programmable amplitude-phase-jointcoding (APJC) information metasurface is proposed and experimentally demonstrated, from which the phase and amplitude of reflected electromagnetic waves can be independently controlled by adjusting the bias voltage of PIN diode integrated in the metaatom. In particular, the reflection amplitude can be continuously controlled from 0.1 to 0.9, and the reflection phase can be switched between two states with about 180 degrees phase difference. Thus, the proposed metasurface is capable of realizing independent 1-bit or multibit amplitude coding and 1-bit phase coding, and both of them can be reprogrammed in real time in broad band from 8 to 13 GHz. The abilities of the programmable APJC information metasurface in manipulating the electromagnetic waves are demonstrated by both numerical simulations and experiments, including to suppress the sidelobes of scattering beam, generate the diffractive waves with arbitrary magnitudes, and so on. These results show unique advantages of APJC information metasurface in real-time independent controls of energy allocation and wavefront tailoring of the electromagnetic waves in a wide frequency band.
This letter studies the information freshness in an industrial Internet of Things (IIoT) network, measured by Age of Information (AoI). We consider a scenario where a sink node collects information update packets from...
详细信息
This letter studies the information freshness in an industrial Internet of Things (IIoT) network, measured by Age of Information (AoI). We consider a scenario where a sink node collects information update packets from different sensors and then uploads the collected packets to an edge server. This scenario has two main requirements: high information freshness (low AoI) and high reliability (low packet error rate, PER). Since update packets are usually short in practice, previous works usually packed and encoded multiple short packets from different sensors into a long packet to improve PER performances. However, while such a joint coding approach improves reliability, it generally leads to longer delay and hence possibly higher AoI. This letter investigates the AoI performance tradeoff by examining the number of packets to be jointly encoded. We consider two AoI metrics, average AoI and bounded AoI. In particular, bounded AoI is the threshold below which the instantaneous AoI falls for a given percentage of the time. Our theoretical analysis and numerical results show that there exist optimal numbers of jointly coded packets that minimize the average AoI and the bounded AoI. Specifically, a smaller number of packets is usually sufficient to achieve both high information freshness and high reliability.
In parallel interference channels, the sum-rate achieved by joint coding among subchannels can exceed the sum of the achievable rate of each subchannel with individual coding. In this paper, a capacity-achieving joint...
详细信息
In parallel interference channels, the sum-rate achieved by joint coding among subchannels can exceed the sum of the achievable rate of each subchannel with individual coding. In this paper, a capacity-achieving joint coding scheme is proposed for parallel symmetric interference channel. First, we provide a motivating example, from which the insights into the joint coding scheme are obtained. Second, we introduce a transmission scheme in two-user parallel symmetric interference channels, where the subchannels can cooperate to cancel interference. Then, by taking advantage of signal level alignment of the interference from different users, we generalize the scheme to multi-user cases. Finally, we prove that our scheme can achieve the sum capacity and illustrate the generalized degrees of freedom gains over individual coding in various interference scenarios.
Non-orthogonal multiple access (NOMA) is a promlsmg candidate for future 5G multiple access. Particularly NOMA uses power domain for multiple access and successive interference canceler (SIC) algorithm for multi-user ...
详细信息
ISBN:
(纸本)9781467390262
Non-orthogonal multiple access (NOMA) is a promlsmg candidate for future 5G multiple access. Particularly NOMA uses power domain for multiple access and successive interference canceler (SIC) algorithm for multi-user detection (MUD). Different from the current downlink transmission scheme for NOMA, a novel joint coding scheme is proposed in this paper and corresponding improved SIC algorithms: zigzag SIC and iterative SIC are designed at the receiver side. Both zigzag SIC and iterative SIC algorithms are based on the novel offset frame structure. The goal is to take advantage of time domain diversity gain and certain long code gain of joint coding and advanced receiver to improve system performance without introduce additional degree of freedom. Practical considerations of NOMA, such as power allocation ratio, SIC, error propagation and channel fading are discussed. Simulation results show the proposed iterative SIC receiver and zigzag SIC receiver can both provide a better performance than conventional SIC receiver, so that the proposed joint coding scheme has obvious advantage over conventional NOMA.
Research on real-time health systems have received great attention during recent years and the needs of high-quality personal multichannel medical signal compression for personal medical product applications are incre...
详细信息
Research on real-time health systems have received great attention during recent years and the needs of high-quality personal multichannel medical signal compression for personal medical product applications are increasing. The international MPEG-4 audio lossless coding (ALS) standard supports a joint channel-coding scheme for improving compression performance of multichannel signals and it is very efficient compression method for multi-channel biosignals. However, the computational complexity of such a multichannel coding scheme is significantly greater than that of other lossless audio encoders. In this paper, we present a multichannel hardware encoder based on a low-complexity joint-coding technique and shared multiplier scheme for portable devices. A joint-coding decision method and a reference channel selection scheme are modified for a low-complexity joint coder. The proposed joint coding decision method determines the optimized joint-coding operation based on the relationship between the cross correlation of residual signals and the compression ratio. The reference channel selection is designed to select a channel for the entropy coding of the joint coding. The hardware encoder operates at a 40 MHz clock frequency and supports two-channel parallel encoding for the multichannel monitoring system. Experimental results show that the compression ratio increases by 0.06%, whereas the computational complexity decreases by 20.72% compared to the MPEG-4 ALS reference software encoder. In addition, the compression ratio increases by about 11.92%, compared to the single channel based bio-signal lossless data compressor.
In this paper, an Internet of things(IoTs) scenario with short packet coding is studied in which a large number of sensors transmit short packets to the edge server through a sink node using a product code. Two main r...
详细信息
ISBN:
(纸本)9781665455442
In this paper, an Internet of things(IoTs) scenario with short packet coding is studied in which a large number of sensors transmit short packets to the edge server through a sink node using a product code. Two main requirements are considered in this scenario which focuses on lower AoI and higher reliability (lower packet error rate, PER). In previous works, short packets are usually encoded into a long packet to improve reliability, which leads to large AoI. In this paper, a product coded packet scheme is considered because the joint packet encoding scheme using product code can be decoded by two ways, decoding of each row/column component code and joint iterative decoding of all component codes. Both of them may lead to a successfully decoding which greatly improves the AoI performance under the same PER conditions. The expression of the average AoI in this IoT system is derived, and the theoretical and numerical results show the relationships between packet length, the number of sensors and AoI performance. Furthermore, the optimal number of jointly-coded packets to minimize the AoI performance also are discussed and illustrated.
Ultra low latency transmission, high bandwidth and ubiquitous access points have greatly alleviated the ＂last mile＂ problem of streaming media live service. However, the complex compression coding of high‐resolutio...
详细信息
Ultra low latency transmission, high bandwidth and ubiquitous access points have greatly alleviated the ＂last mile＂ problem of streaming media live service. However, the complex compression coding of high‐resolution video is still inevitable before transmission. We combine the YOLOv5 target detection algorithm with the video encoder, segment the foreground and background of the video to be encoded with the help of the improved YOLOv5, and then encode the foreground and background respectively. The foreground area of the video uses the normal motion search to find the motion vector, and the background area uses the global motion parameters to describe its overall motion. This scheme can get the target detection results and encode the video at the same time. The experimental results show that the proposed scheme greatly reduces the complexity of the encoder with less loss of video quality, up to 82%. At the same time, the target detection task is completed.
Embedded zero tree wavelet coding (EZW) is an effective image coding algorithm, Study on the output stream of P, N, T, Z symbol, they are appeared with inequality probability. In order to reduce the desired number of ...
详细信息
ISBN:
(纸本)9781479941711
Embedded zero tree wavelet coding (EZW) is an effective image coding algorithm, Study on the output stream of P, N, T, Z symbol, they are appeared with inequality probability. In order to reduce the desired number of digits used in coding, this paper proposes an EZW and Huffman joint encoding algorithm. The experimental results shows, compared with the independent EZW algorithm, the joint encoding algorithm can improve the efficiency of image compression and coding.
To achieve high-efficiency and high-reliable data transfer between each pair of users in multiinput multi-output(MIMO) wireless communication,researchers show concerns for communication system that has become a key te...
详细信息
To achieve high-efficiency and high-reliable data transfer between each pair of users in multiinput multi-output(MIMO) wireless communication,researchers show concerns for communication system that has become a key technology,which attracts many both at home and *** this article,a coding scheme that can achieve full-rate transmission and full-diversity bit error rate(BER) performance is proposed for the network model that contains two pairs of users,and each user is fitted to four *** this scheme,the orthogonal properties of Alamouti space-time code are used for alignment interference at the receiver and the joint coding technology is used to get the full-diversity coding gain *** the simulation results confirm the validity and correctness of the proposed coding transmission scheme.
暂无评论