Video compression is a basic requirement for consumer and professional video applications alike. Video coding standards such as H.264/AVC and H.265/HEVC are widely deployed in the market to enable efficient use of bandwidth and storage for many video applications. To reduce the coding artifacts and improve the compression efficiency, neural network based loop filtering of the reconstructed video has been developed in the literature. However, loop filtering is a challenging task due to the variation in video content and sampling densities. In this paper, we propose a on-line scaling based multi-density attention network for loop filtering in video compression. The core of our approach lies in several aspects: (a) parallel multi-resolution convolution streams for extracting multi-density features, (b) single attention branch to learn the sample correlations and generate mask maps, (c) a channel-mutual attention procedure to fuse the data from multiple branches, (d) on-line scaling technique to further optimize the output results of network according to the actual signal. The proposed multi-density attention network learns rich features from multiple sampling densities and performs robustly on video content of different resolutions. Moreover, the online scaling process enhances the signal adaptability of the off-line pre-trained model. Experimental results show that 10.18% bit-rate reduction at the same video quality can be achieved over the latest Versatile Video Coding (VVC) standard. The objective performance of the proposed algorithm outperforms the state-of-the-art methods and the subjective quality improvement is obvious in terms of detail preservation and artifact alleviation.
This paper introduces a novel framework for end-to-end learned video coding. Image compression is generalized through conditional coding to exploit information from reference frames, allowing to process intra and inter frames with the same coder. The system is trained through the minimization of a rate-distortion cost, with no pre-training or proxy loss. Its flexibility is assessed under three coding configurations (All Intra, Low-delay P and Random Access), where it is shown to achieve performance competitive with the state-of-the-art video codec HEVC.
The framework of dominant learned video compression methods is usually composed of motion prediction modules as well as motion vector and residual image compression modules, suffering from its complex structure and error propagation problem. Approaches have been proposed to reduce the complexity by replacing motion prediction modules with implicit flow networks. Error propagation aware training strategy is also proposed to alleviate incremental reconstruction errors from previously decoded frames. Although these methods have brought some improvement, little attention has been paid to the framework itself. Inspired by the success of learned image compression through simplifying the framework with a single deep neural network, it is natural to expect a better performance in video compression via a simple yet appropriate framework. Therefore, we propose a framework to directly compress raw-pixel frames (rather than residual images), where no extra motion prediction module is required. Instead, an entropy model is used to estimate the spatiotemporal redundancy in a latent space rather than pixel level, which significantly reduces the complexity of the framework. Specifically, the whole framework is a compression module, consisting of a unified auto-encoder which produces identically distributed latents for all frames, and a spatiotemporal entropy estimation model to minimize the entropy of these latents. Experiments showed that the proposed method outperforms state-of-the-art (SOTA) performance under the metric of multiscale structural similarity (MS-SSIM) and achieves competitive results under the metric of PSNR.
Video compression is a critical component of Internet video delivery. Recent work has shown that deep learning techniques can rival or outperform human-designed algorithms, but these methods are significantly less compute and power-efficient than existing codecs. This paper presents a new approach that augments existing codecs with a small, content-adaptive super-resolution model that significantly boosts video quality. Our method, SRVC, encodes video into two bitstreams: (i) a content stream, produced by compressing downsampled low-resolution video with the existing codec, (ii) a model stream, which encodes periodic updates to a lightweight super-resolution neural network customized for short segments of the video. SRVC decodes the video by passing the decompressed low-resolution video frames through the (time-varying) super-resolution model to reconstruct high-resolution video frames. Our results show that to achieve the same PSNR, SRVC requires 16% of the bits-per-pixel of H.265 in slow mode, and 2% of the bits-per-pixel of DVC, a recent deep learning-based video compression scheme. SRVC runs at 90 frames per second on a NVIDIA V100 GPU.
We generalize the 'bits back with ANS' method to time-series models with a latent Markov structure. This family of models includes hidden Markov models (HMMs), linear Gaussian state space models (LGSSMs) and many more. We provide experimental evidence that our method is effective for small scale models, and discuss its applicability to larger scale settings such as video compression.
Temporal interpolation has the potential to be a powerful tool for video compression. Existing methods for frame interpolation do not discriminate between video textures and generally invoke a single general model capable of interpolating a wide range of video content. However, past work on video texture analysis and synthesis has shown that different textures exhibit vastly different motion characteristics and they can be divided into three classes (static, dynamic continuous and dynamic discrete). In this work, we study the impact of video textures on video frame interpolation, and propose a novel framework where, given an interpolation algorithm, separate models are trained on different textures. Our study shows that video texture has significant impact on the performance of frame interpolation models and it is beneficial to have separate models specifically adapted to these texture classes, instead of training a single model that tries to learn generic motion. Our results demonstrate that models fine-tuned using our framework achieve, on average, a 0.3dB gain in PSNR on the test set used.
As an important perceptual characteristic of the Human Visual System (HVS), the Just Noticeable Difference (JND) has been studied for decades with image and video processing (e.g., perceptual visual signal compression). However, there is little exploration on the existence of JND for the Deep Machine Vision (DMV), although the DMV has made great strides in many machine vision tasks. In this paper, we take an initial attempt, and demonstrate that the DMV has the JND, termed as the DMV-JND. We then propose a JND model for the image classification task in the DMV. It has been discovered that the DMV can tolerate distorted images with average PSNR of only 9.56dB (the lower the better), by generating JND via unsupervised learning with the proposed DMV-JND-NET. In particular, a semantic-guided redundancy assessment strategy is designed to restrain the magnitude and spatial distribution of the DMV-JND. Experimental results on image classification demonstrate that we successfully find the JND for deep machine vision. Our DMV-JND facilitates a possible direction for DMV-oriented image and video compression, watermarking, quality assessment, deep neural network security, and so on.
Although the video compression ratio nowadays becomes higher, the video coders such as H.264/AVC, H.265/HEVC, H.266/VVC always suffer from the video artifacts. In this paper, we design a neural network to enhance the quality of the compressed frame by leveraging the block information, called B-DRRN (Deep Recursive Residual Network with Block information). Firstly, an extra network branch is designed for leveraging the block information of the coding unit (CU). Moreover, to avoid a great increase in the network size, Recursive Residual structure and sharing weight techniques are applied. We also conduct a new large-scale dataset with 209,152 training samples. Experimental results show that the proposed B-DRRN can reduce 6.16% BD-rate compared to the HEVC standard. After efficiently adding an extra network branch, this work can improve the performance of the main network without increasing any memory for storing.
Rate-Distortion Optimized Quantization (RDOQ) has played an important role in the coding performance of recent video compression standards such as H.264/AVC, H.265/HEVC, VP9 and AV1. This scheme yields significant reductions in bit-rate at the expense of relatively small increases in distortion. Typically, RDOQ algorithms are prohibitively expensive to implement on real-time hardware encoders due to their sequential nature and their need to frequently obtain entropy coding costs. This work addresses this limitation using a neural network-based approach, which learns to trade-off rate and distortion during offline supervised training. As these networks are based solely on standard arithmetic operations that can be executed on existing neural network hardware, no additional area-on-chip needs to be reserved for dedicated RDOQ circuitry. We train two classes of neural networks, a fully-convolutional network and an auto-regressive network, and evaluate each as a post-quantization step designed to refine cheap quantization schemes such as scalar quantization (SQ). Both network architectures are designed to have a low computational overhead. After training they are integrated into the HM 16.20 implementation of HEVC, and their video coding performance is evaluated on a subset of the H.266/VVC SDR common test sequences. Comparisons are made to RDOQ and SQ implementations in HM 16.20. Our method achieves 1.64% BD-rate savings on luminosity compared to the HM SQ anchor, and on average reaches 45% of the performance of the iterative HM RDOQ algorithm.
Compression methods based on inpainting are an evolving alternative to classical transform-based codecs for still images. Attempts to apply these ideas to video compression are rare, since reaching real-time performance is very challenging. Therefore, current approaches focus on simplified frame-by-frame reconstructions that ignore temporal redundancies. As a remedy, we propose a highly efficient, real-time capable prediction and correction approach that fully relies on partial differential equations (PDEs) in all steps of the codec: Dense variational optic flow fields yield accurate motion-compensated predictions, while homogeneous diffusion inpainting is applied for intra prediction. To compress residuals, we introduce a new highly efficient block-based variant of pseudodifferential inpainting. Our novel architecture outperforms other inpainting-based video codecs in terms of both quality and speed. For the first time in inpainting-based video compression, we can decompress FullHD (1080p) videos in real-time with a fully CPU-based implementation, outperforming previous approaches by roughly one order of magnitude.
The AV1 video compression format is developed by the Alliance for Open Media consortium. It achieves more than 30% reduction in bit-rate compared to its predecessor VP9 for the same decoded video quality. This paper provides a technical overview of the AV1 codec design that enables the compression performance gains with considerations for hardware feasibility.
Motion estimation is a key component of any modern video codec. Our understanding of motion and the estimation of motion from video has come a very long way since 2000. More than 135 different algorithms have been recently reviewed by Scharstein et al http://vision.middlebury.edu/flow/. These new algorithms differ markedly from Block Matching which has been the mainstay of video compression for some time. This paper presents comparisons of H.264 and MP4 compression using different motion estimation methods. In so doing we present as well methods for adapting pre-computed motion fields for use within a codec. We do not observe significant gains to be had with the methods chosen w.r.t. Rate Distortion tradeoffs but the results reflect a significantly more complex interrelationship between motion and compression than would be expected. There remains much more to be done to improve the coverage of this comparison to the emerging standards but these initial results show that there is value in these explorations.
The video technology scenery has been very vivid over the past years, with novel video coding technologies introduced that promise improved compression performance over state-of-the-art technologies. Despite the fact that a lot of video datasets are available, representative content of the wide parameter space along with subjective evaluations of variations of encoded content from an unpartial end is required. In response to this requirement, this paper features a dataset, the BVI-CC. Three video codecs were deployed to create the variations of the encoded sequences: High Efficiency Video Coding (HEVC) Test Model (HM), AOMedia Video 1 (AV1), and Versatile Video Coding (VVC) Test Model (VTM). Nine source video sequences were carefully selected to offer both diversity and representativeness in the spatio-temporal domain. Different spatial resolution versions of the sequences were created and encoded by all three codecs at pre-defined target bit rates. The compression efficiency of the codecs was evaluated with commonly used objective quality metrics, and the subjective quality of their reconstructed content was also evaluated through psychophysical experiments. Additionally, an adaptive bit rate (convex hull rate-distortion optimization across spatial resolutions) test case was assessed using both objective and subjective evaluations. Finally, the computational complexities of the tested codecs were examined. All data have been made publicly available as part of the dataset, which can be used for coding performance evaluation and video quality metric development.
The pursuit of higher compression efficiency continuously drives the advances of video coding technologies. Fundamentally, we wish to find better "predictions" or "priors" that are reconstructed previously to remove the signal dependency efficiently and to accurately model the signal distribution for entropy coding. In this work, we propose a neural video compression framework, leveraging the spatial and temporal priors, independently and jointly to exploit the correlations in intra texture, optical flow based temporal motion and residuals. Spatial priors are generated using downscaled low-resolution features, while temporal priors (from previous reference frames and residuals) are captured using a convolutional neural network based long-short term memory (ConvLSTM) structure in a temporal recurrent fashion. All of these parts are connected and trained jointly towards the optimal rate-distortion performance. Compared with the High-Efficiency Video Coding (HEVC) Main Profile (MP), our method has demonstrated averaged 38% Bjontegaard-Delta Rate (BD-Rate) improvement using standard common test sequences, where the distortion is multi-scale structural similarity (MS-SSIM).
The existing video coding standards such as H.264/AVC and High Efficiency Video Coding (HEVC) have been designed based on the statistical properties of Low Dynamic Range (LDR) videos and are not accustomed to the characteristics of High Dynamic Range (HDR) content. In this study, we investigate the performance of the latest LDR video compression standard, HEVC, as well as the recent widely commercially used video compression standard, H.264/AVC, on HDR content. Subjective evaluations of results on an HDR display show that viewers clearly prefer the videos coded via an HEVC-based encoder to the ones encoded using an H.264/AVC encoder. In particular, HEVC outperforms H.264/AVC by an average of 10.18% in terms of mean opinion score and 25.08% in terms of bit rate savings.
While there exists a wide variety of Low Dynamic Range (LDR) quality metrics, only a limited number of metrics are designed specifically for the High Dynamic Range (HDR) content. With the introduction of HDR video compression standardization effort by international standardization bodies, the need for an efficient video quality metric for HDR applications has become more pronounced. The objective of this study is to compare the performance of the existing full-reference LDR and HDR video quality metrics on HDR content and identify the most effective one for HDR applications. To this end, a new HDR video dataset is created, which consists of representative indoor and outdoor video sequences with different brightness, motion levels and different representing types of distortions. The quality of each distorted video in this dataset is evaluated both subjectively and objectively. The correlation between the subjective and objective results confirm that VIF quality metric outperforms all to ther tested metrics in the presence of the tested types of distortions.
As the evolution of multiview display technology is bringing glasses-free 3DTV closer to reality, MPEG and VCEG are preparing an extension to HEVC to encode multiview video content. View synthesis in the current version of the 3D video codec is performed using PSNR as a quality metric measure. In this paper, we propose a full- reference Human-Visual-System based 3D video quality metric to be used in multiview encoding as an alternative to PSNR. Performance of our metric is tested in a 2-view case scenario. The quality of the compressed stereo pair, formed from a decoded view and a synthesized view, is evaluated at the encoder side. The performance is verified through a series of subjective tests and compared with that of PSNR, SSIM, MS-SSIM, VIFp, and VQM metrics. Experimental results showed that our 3D quality metric has the highest correlation with Mean Opinion Scores (MOS) compared to the other tested metrics.
In the popular video coding trend, the encoder has the task to exploit both spatial and temporal redundancies present in the video sequence, which is a complex procedure. As a result almost all video encoders have five to ten times more complexity than their decoders. In a video compression process, one of the main tasks at the encoder side is motion estimation which is to extract the temporal correlation between frames. Distributed video coding (DVC) proposed the idea that can lead to low complexity encoders and higher complexity decoders. DVC is a new paradigm in video compression based on the information theoretic ideas of Slepian-Wolf and Wyner-Ziv theorems. Wyner-Ziv coding is naturally robust against transmission errors and can be used for joint source and channel coding. Side Information is one of the key components of the Wyner-Ziv decoder. Better side information generation will result in better functionality of Wyner-Ziv coder. In this paper we proposed a new method that can generate side information with a better quality and thus better compression. We have used HVS (human visual system) based image quality metrics as our quality criterion. The motion estimation we used in the decoder is modified due to these metrics such that we could obtain finer side information. The motion compensation is optimized for perceptual quality metrics and leads to better side information generation compared to con- ventional MSE (mean squared error) or SAD (sum of absolute difference) based motion compensation currently used in the literature. Better motion compensation means better compression.
In this work we propose a novel deep learning approach for ultra-low bitrate video compression for video conferencing applications. To address the shortcomings of current video compression paradigms when the available bandwidth is extremely limited, we adopt a model-based approach that employs deep neural networks to encode motion information as keypoint displacement and reconstruct the video signal at the decoder side. The overall system is trained in an end-to-end fashion minimizing a reconstruction error on the encoder output. Objective and subjective quality evaluation experiments demonstrate that the proposed approach provides an average bitrate reduction for the same visual quality of more than 80% compared to HEVC.
Inspired by the facts that retinal cells actually segregate the visual scene into different attributes (e.g., spatial details, temporal motion) for respective neuronal processing, we propose to first decompose the input video into respective spatial texture frames (STF) at its native spatial resolution that preserve the rich spatial details, and the other temporal motion frames (TMF) at a lower spatial resolution that retain the motion smoothness; then compress them together using any popular video coder; and finally synthesize decoded STFs and TMFs for high-fidelity video reconstruction at the same resolution as its native input. This work simply applies the bicubic resampling in decomposition and HEVC compliant codec in compression, and puts the focus on the synthesis part. For resolution-adaptive synthesis, a motion compensation network (MCN) is devised on TMFs to efficiently align and aggregate temporal motion features that will be jointly processed with corresponding STFs using a non-local texture transfer network (NL-TTN) to better augment spatial details, by which the compression and resolution resampling noises can be effectively alleviated with better rate-distortion efficiency. Such "Decomposition, Compression, Synthesis (DCS)" based scheme is codec agnostic, currently exemplifying averaged $\approx$1 dB PSNR gain or $\approx$25% BD-rate saving, against the HEVC anchor using reference software. In addition, experimental comparisons to the state-of-the-art methods and ablation studies are conducted to further report the efficiency and generalization of DCS algorithm, promising an encouraging direction for future video coding.