Deep learning-based complex hologram compression enhanced by ringing reduction

Computational imaging and light field technology promise to deliver the required six-degrees-of-freedom for natural scenes in virtual reality. Already existing extensions of standardized video coding formats, such as multi-view coding and multi-view plus depth, are the most conventional light field video coding solutions at the moment. The latest multi-view coding format, which is a direct extension of the high efficiency video coding (HEVC) standard, is called multi-view HEVC (or MV-HEVC). MV-HEVC treats each light field view as a separate video sequence, and uses syntax elements similar to standard HEVC for exploiting redundancies between neighboring views. To achieve this, inter-view and temporal prediction schemes are deployed with the aim to find the most optimal trade-off between coding performance and reconstruction quality. The number of possible prediction structures is unlimited and many of them are proposed in the literature. Although some of them are efficient in terms of compression ratio, they complicate random access due to the dependencies on previously decoded pixels or frames. Random access is an important feature in video delivery, and a crucial requirement in multi-view video coding. In this work, we propose and compare different prediction structures for coding light field video using MV-HEVC with a focus on both compression efficiency and random accessibility. Experiments on three different short-baseline light field video sequences show the trade-off between bit-rate and distortion, as well as the average number of decoded views/frames, necessary for displaying any random frame at any time instance. The findings of this work indicate the most appropriate prediction structure depending on the available bandwidth and the required degree of random access.

In order to realize copyright protection and piracy tracking for ever-increasing videos, video information hiding has gained more and more attention. For the latest video coding standard — High Efficiency Video Coding (HEVC), most of the information hiding algorithms are based on the same characteristics as those in MPEG4 and H.264, such as DCT coefficients, intra prediction modules and motion vectors. However, few algorithms are reported to hide information based on the innovation modules introduced by HEVC. This paper has proposed a high capacity and multilevel information hiding algorithm based on PU partition modes in P-frames, which is considered to be one of the most important innovative features of HEVC. The proposed algorithm consists of two stages: first round calculation and modification process. The first round calculation is to record the PU partition modes selected by HEVC. The modification process is to force each of PU mode into one of the encoded groups according to the embedding binary bits, and then use the modified PU partition modes for the left HEVC encoding process. The experimental results show that the proposed algorithm can achieve high capacity and visual quality, and low bit rate increase, which significantly outperforms state-of-the-art works in embedding capacity.

High quality videos became an essential requirement in recent applications. High efficiency video coding (HEVC) standard provides an efficient solution for high quality videos at lower bit rates. On the other hand, HEVC comes with much higher computational cost. In particular, motion estimation (ME) in HEVC, consumes the largest amount of computations. Therefore, fast ME algorithms and hardware accelerators are proposed in order to speed‐up integer ME in HEVC. This study presents a fast centre search algorithm (FCSA) and an adaptive search window algorithm (ASWA) for integer pixel ME in HEVC. In addition, centre adaptive search algorithm, a combination of the two proposed algorithms FCSA and ASWA, is proposed in order to achieve the best performance. Experimental results show notable speed‐up in terms of encoding time and bit rate saving with tolerable peak signal‐to‐noise ratio (PSNR) quality degradation. The proposed fast search algorithms reduce the computational complexity of the HEVC encoder by 57%. This improvement is accompanied with a modest average PSNR loss of 0.014 dB and an increase by 0.6385% in terms of bit rate when compared with related works.

In this brief, a new very-large-scale integrated (VLSI) integer transform architecture is proposed for the High Efficiency Video Coding (HEVC) encoder. The architecture is designed based on the signed bit-plane transform (SBT) matrices, which are derived from the bit-plane decompositions of the integer transform matrices in HEVC. Mathematically, an integer transform matrix can be equally expressed by the binary weighted sum of several SBT matrices that are only composed of binary 0 or ±1. The SBT matrices are very simple and have lower bit width than the original integer transform in the form. The SBT matrices are also sparse and there are many zero elements. The sparse characteristic of SBT matrices is very helpful for saving the addition operators of SBT. In the proposed architecture, instead of the original integer transform in high bit width, the video data can be respectively transformed with the SBT matrices in lower bit width. As a result, the delay of the transform unit circuit can be significantly reduced with the proposed SBT. Moreover, exploiting the redundant element characteristic of SBT matrices, in which the elements are 0 or ±1, the adder reuse strategy is proposed for our transform architecture, which can save the circuit area efficiently. The simulation results show that by employing the proposed strategies the VLSI transform architecture can be synthesized in a proper area with a high working frequency and low latency. The architecture can support all HEVC encoders coding ultra high-definition video sequences in real time.

High Efficient Video Coding (HEVC) is the latest standard in the video transmission sector, and is designed to improve the compression rate and efficiency of video streaming applications. However, one of the many design constraints to improve HEVC is the reliability of transmission. HEVC introduces a sever output degradation when error propagation occurs because of its high compression rate. In this paper, the performance of HEVC is evaluated by video streaming over a Rayleigh distributed environment. The transmission over Rayleigh distributed environment will generate lost and dropped packets because of attenuation in the reception power and limited capacity of the network. The packet loss is then analyzed such that it obeys the statistical envelope of the propagation model used. The wireless simulation generates error patterns which depicts which frame and slice IDs in the video sequence have experienced a packet loss, which will be used as reference by the modified HEVC decoder for motion concealment. The results measured from the decoded video quality are calculated using the peak-signal-to-noise ratio (PSNR) which is compared to the percent error rate. These results were obtained by setting bit rates of 32 kbps and 128 kbps for three video sequences with different motion levels, as well as setting the bytes per slice to 16. In a wireless network with a Rayleigh distributed channel, increasing the bit rate and motion level of the video sequence will decrease the overall PSNR.

In this paper, we present a pixel-wise unified rate quantization (R-Q) model for a low-complexity rate control on configurable coding units of high efficiency video coding (HEVC). In the case of HEVC, which employs hierarchical coding block structure, multiple R-Q models can be employed for the various block sizes. However, we found that the ratios of distortions over bits for all the blocks are a nearly constant because of employment of the rate distortion optimization technique. Hence, one relationship model between rate and quantization can be derived from the characteristic of similar ratios of distortions over bits regardless of block sizes. Thus, we propose the pixel-wise unified R-Q model for HEVC rate control working on the multi-level for all block sizes. We employ a simple leaky bucket model for bit control. The rate control based on the proposed pixel-wise unified R-Q model is implemented on HEVC test model 6.1 (HM6.1). According to the evaluation for the proposed rate control, the average matching percentage to target bitrates is 99.47% and the average PSNR degradation is 0.76 dB. Based on the comparative study, we found that the proposed rate control shows low bit fluctuation and good RD performance, compared to R-lambda rate control for long sequences.

Nowadays, image compression is very important for efficient data storage and transmission. This paper presents a dictionary-based intra-prediction framework for image compression using sparse representation, with the construction of trained over-complete dictionaries. The intra-prediction residuals selected from different images and K-SVD algorithm are used to train over-complete dictionaries. The trained dictionaries are integrated into the intra-prediction framework for efficient image compression. In this proposed method, first, intra-prediction is applied over an image and then prediction residuals of the image are encoded using sparse representation. Sparse approximation algorithm and trained dictionaries are employed for encoding of prediction residuals of the image. The coefficients obtained from sparse representation are used for encoding. For efficient sparse representation with fewer dictionary coefficients, an adaptive sparse image partitioning method is introduced. Simulation result demonstrates that the proposed image compression method yields improved encoding efficiency as compared to existing schemes.

Transform/quantization (T/Q) is one of the high complex modules in high efficiency video coding (HEVC). In this paper, all zero block (AZB) detection algorithm for HEVC is proposed to reduce the complexity of T/Q. The proposed AZB detection algorithm is based on the quantization level of the maximum transform coefficient, which is suitable for both uniform quantizer (UQ) and rate-distortion optimized quantizer (RDOQ). Experimental results show that 46% complexity of T/Q for UQ and 42% for RDOQ can be reduced with negligible loss of video quality and compression efficiency, outperforming the state-of-the-art methods.

New video coding solutions, such as the HEVC (High Efficiency Video Coding) standard being developed by JCT-VC (Joint Collaborative Team on Video Coding), are typically designed for high resolution video content. Increasing video resolution creates two basic requirements for practical video codecs; those need to be able to provide compression efficiency superior to prior video coding solutions and the computational requirements need to be aligned with the foreseeable hardware platforms. This paper proposes an intra prediction method which is designed to provide high compression efficiency and which can be implemented effectively in resource constrained environments making it applicable to wide range of use cases. When designing the method, special attention was given to the algorithmic definition of the prediction sample generation, in order to be able to utilize the same reconstruction process at different block sizes. The proposed method outperforms earlier variations of the same family of technologies significantly and consistently across different classes of video material, and has recently been adopted as the directional intra prediction method for the draft HEVC standard. Experimental results show that the proposed method outperforms the H.264/AVC intra prediction approach on average by 4.8 %. For sequences with dominant directional structures, the coding efficiency gains become more significant and exceed 10 %.

High efficiency Video Coding (HEVC) is the state-of-the-art video coding technique capable of encoding Ultra High Definition (UHD) videos with better compression efficiency and has better reconstruction quality for the same bitrate as compared to its predecessors. Better compression is possible due to its complex partition and prediction methods. These benefits are at the cost of increased computational complexity, which in turn increases resource consumption and processing time. In this work, we design and implement three different architectures, viz: 1) Fully Sequential Architecture (FSA), 2) Semi-parallel Architecture (SPA) and 3) Fully Parallel Architecture (FPA), for the Intra prediction of HEVC on Field Programmable Gate Arrays (FPGA) and discuss the results. These three configurations are tested for the prediction units of sizes 4×4, 8×8 and 16×16. Results show that FSA uses nearly 70% fewer resources than FPA. Also FSA uses 51.73%, 54.33% and 52.2% less resources than SPA for 4×4, 8×8 and 16×16 block sizes, respectively. Also, the FPA implemented for all three pediction unit (PU) sizes is nearly 22 times and 5 times faster than the FSA and SPA, respectively.

In order to satisfy the requirement of bandwidth and storage capacity, high efficient image compression coding method is one of the key technologies. The general image compression methods only encode the original pixels without any analysis. A deep space image compression algorithm based on the region of interest (ROI) is proposed in the paper. For deep space exploration, only parts of the image are interested in depending on the application background. Some image area such as secondary planet, star and satellite can be considered as ROI. The proposed method includes image segmentation and different image compressions for different regions. The algorithm is characterized with higher image signal noise ratio (ISNR) of the reconstructed image and lower computation complexity, and the image detail preserving capability of the algorithm is better than that of JPEG2000. Because of its simplicity, fastness, and small storage, the algorithm is easy to be realized in hardware and suitable for space borne application.

Successive image compression refers to the process of repeated encoding and decoding of an image. It frequently occurs during sharing, manipulation, and re-distribution of images. While deep learning-based methods have made significant progress for single-step compression, thorough analysis of their performance under successive compression has not been conducted. In this paper, we conduct comprehensive analysis of successive deep image compression. First, we introduce a new observation, instability of successive deep image compression, which is not observed in JPEG, and discuss causes of the instability. Then, we conduct a successive image compression benchmark for the state-of-the-art deep learning-based methods, and analyze the factors that affect the instability in a comparative manner. Finally, we propose a new loss function for training deep compression models, called feature identity loss, to mitigate the instability of successive deep image compression.

Abstract. The Lo's modified rescaled adjusted range test (R/S test) (Lo, 1991), GPH test (Geweke and Porter-Hudak, 1983) and two approximate maximum likelihood estimation methods, i.e., Whittle's estimator (W-MLE) and another one implemented in S-Plus (S-MLE) based on the algorithm of Haslett and Raftery (1989) are evaluated through intensive Monte Carlo simulations for detecting the existence of long-memory. It is shown that it is difficult to find an appropriate lag q for Lo's test for different short-memory autoregressive (AR) and fractionally integrated autoregressive and moving average (ARFIMA) processes, which makes the use of Lo's test very tricky. In general, the GPH test outperforms the Lo's test, but for cases where a strong short-range dependence exists (e.g., AR(1) processes with φ=0.95 or even 0.99), the GPH test gets useless, even for time series of large data size. On the other hand, the estimates of d given by S-MLE and W-MLE seem to give a good indication of whether or not the long-memory is present. The simulation results show that data size has a significant impact on the power of all the four methods because the availability of larger samples allows one to inspect the asymptotical properties better. Generally, the power of Lo's test and GPH test increases with increasing data size, and the estimates of d with GPH method, S-MLE method and W-MLE method converge with increasing data size. If no large enough data set is available, we should be aware of the possible bias of the estimates. The four methods are applied to daily average discharge series recorded at 31 gauging stations with different drainage areas in eight river basins in Europe, Canada and USA to detect the existence of long-memory. The results show that the presence of long-memory in 29 daily series is confirmed by at least three methods, whereas the other two series are indicated to be long-memory processes with two methods. The intensity of long-memory in daily streamflow processes has only a very weak positive relationship with the scale of watershed.

High Efficiency Video Coding (HEVC) is a new standard under development by the ISO and ITU-T. In this work an encoder which incorporates HVS models into the basic coding structure of the recommended HEVC standard is implemented, to improve the quality of the standard coder. Since human eye ultimately decides the quality of the received signal and the sole aim of HEVC is coding efficiency and quality improvement, this can be best achieved by selecting and incorporating appropriate Human Visual System (HVS) models into the standard coding structure. The proposed coder incorporates adaptive HVS based quantization matrices to improve the quality of reconstruction. The performance of the coder is compared with that of standard codecs such as MPEG-2, H.264/AVC and HEVC without perceptual models. The performance improvement for the proposed coder is evaluated using various objective quality measures such as PSNR, MSE, MSSIM, VIF, and VSNR.

The recent introduction of the High Efficiency Video Coding (HEVC) standard provides opportunities to improve medical image compression in picture archiving and communications systems. In this paper, we propose improvements to the HEVC intra coding process for lossless compression of grayscale anatomical medical images, which are characterized by their large amount of edges. Specifically, we propose alternative angular and planar prediction modes that are based on sample-wise differential pulse code modulation (DPCM) with an increased range of directionalities. We also propose an implementation of the DPCM decoding process that maintains the block-wise coding structure of HEVC. Evaluation results on various medical images show that the proposed DPCM modes efficiently predict the large amount of edges in these images achieving bit-rate savings of up to 15%.