arXiv subjects

Mai Xu

Publications and source records attributed to Mai Xu.

61 records · Page 4Linked to original sources

arXiv2018

Reducing Complexity of HEVC: A Deep Learning Approach

High Efficiency Video Coding (HEVC) significantly reduces bit-rates over the proceeding H.264 standard but at the expense of extremely high encoding complexity. In HEVC, the quad-tree partition of coding unit (CU) consumes a large proportion of the HEVC encoding complexity, due to the bruteforce search for rate-distortion optimization (RDO). Therefore, this paper proposes a deep learning approach to predict the CU partition for reducing the HEVC complexity at both intra- and inter-modes, which is based on convolutional neural network (CNN) and long- and short-term memory (LSTM) network. First, we establish a large-scale database including substantial CU partition data for HEVC intra- and inter-modes. This enables deep learning on the CU partition. Second, we represent the CU partition of an entire coding tree unit (CTU) in the form of a hierarchical CU partition map (HCPM). Then, we propose an early-terminated hierarchical CNN (ETH-CNN) for learning to predict the HCPM. Consequently, the encoding complexity of intra-mode HEVC can be drastically reduced by replacing the brute-force search with ETH-CNN to decide the CU partition. Third, an early-terminated hierarchical LSTM (ETH-LSTM) is proposed to learn the temporal correlation of the CU partition. Then, we combine ETH-LSTM and ETH-CNN to predict the CU partition for reducing the HEVC complexity for inter-mode. Finally, experimental results show that our approach outperforms other state-of-the-art approaches in reducing the HEVC complexity at both intra- and inter-modes.

cs.CV↗

arXiv2018

Multi-Frame Quality Enhancement for Compressed Video

The past few years have witnessed great success in applying deep learning to enhance the quality of compressed image/video. The existing approaches mainly focus on enhancing the quality of a single frame, ignoring the similarity between consecutive frames. In this paper, we investigate that heavy quality fluctuation exists across compressed video frames, and thus low quality frames can be enhanced using the neighboring high quality frames, seen as Multi-Frame Quality Enhancement (MFQE). Accordingly, this paper proposes an MFQE approach for compressed video, as a first attempt in this direction. In our approach, we firstly develop a Support Vector Machine (SVM) based detector to locate Peak Quality Frames (PQFs) in compressed video. Then, a novel Multi-Frame Convolutional Neural Network (MF-CNN) is designed to enhance the quality of compressed video, in which the non-PQF and its nearest two PQFs are as the input. The MF-CNN compensates motion between the non-PQF and PQFs through the Motion Compensation subnet (MC-subnet). Subsequently, the Quality Enhancement subnet (QE-subnet) reduces compression artifacts of the non-PQF with the help of its nearest PQFs. Finally, the experiments validate the effectiveness and generality of our MFQE approach in advancing the state-of-the-art quality enhancement of compressed video. The code of our MFQE approach is available at https://github.com/ryangBUAA/MFQE.git

cs.CV↗

arXiv2018

Saliency-Guided Complexity Control for HEVC Decoding

The latest High Efficiency Video Coding (HEVC) standard significantly improves coding efficiency over its previous video coding standards. The expense of such improvement is enormous computational complexity, from both encoding and decoding sides. Since computational capability and power capacity are diverse across portable devices, it is necessary to reduce decoding complexity to a target with tolerable quality loss, so called complexity control. This paper proposes a Saliency-Guided Complexity Control (SGCC) approach for HEVC decoding, which reduces the decoding complexity to the target with minimal perceptual quality loss. First, we establish the SGCC formulation to minimize perceptual quality loss at the constraint on reduced decoding complexity, which is achieved via disabling Deblocking Filter (DF) and simplifying Motion Compensation (MC) of some non-salient Coding Tree Units (CTUs). One important component in this formulation is the modelled relationship between decoding complexity reduction and DF disabling/MC simplification, which determines the control accuracy of our approach. Another component is the modelled relationship between quality loss and DF disabling/MC simplification, responsible for optimizing perceptual quality. By solving the SGCC formulation for a given target complexity, we can obtain the DF and MC settings of each CTU, and then decoding complexity can be reduced to the target. Finally, the experimental results validate the effectiveness of our SGCC approach, from the aspects of control performance, complexity-distortion performance, fluctuation of quality loss and subjective quality.

cs.MM↗

arXiv2017

Watching Videos with Certain and Constant Quality: PID-based Quality Control Method

In video coding, compressed videos with certain and constant quality can ensure quality of experience (QoE). To this end, we propose in this paper a novel PID-based quality control (PQC) method for video coding. Specifically, a formulation is modelled to control quality of video coding with two objectives: minimizing control error and quality fluctuation. Then, we apply the Laplace domain analysis to model the relationship between quantization parameter (QP) and control error in this formulation. Given the relationship between QP and control error, we propose a solution to the PQC formulation, such that videos can be compressed at certain and constant quality. Finally, experimental results show that our PQC method is effective in both control accuracy and quality fluctuation.

cs.MM↗

arXiv2016

Secure Multicast Communications with Private Jammers

This paper investigates secrecy rate optimization for a multicasting network, in which a transmitter broadcasts the same information to multiple legitimate users in the presence of multiple eavesdroppers. In order to improve the achievable secrecy rates, private jammers are employed to generate interference to confuse the eavesdroppers. These private jammers charge the legitimate transmitter for their jamming services based on the amount of interference received at the eavesdroppers. Therefore, this secrecy rate maximization problem is formulated as a Stackelberg game, in which the private jammers and the transmitter are the leaders and the follower of the game, respectively. A fixed interference price scenario is considered first, in which a closed-form solution is derived for the optimal amount of interference generated by the jammers to maximize the revenue of the legitimate transmitter. Based on this solution, the Stackelberg equilibrium of the proposed game, at which both legitimate transmitter and the private jammers achieve their maximum revenues, is then derived. Simulation results are also provided to validate these theoretical derivations.

cs.GT↗

arXiv2012

Rateless Codes with Progressive Recovery for Layered Multimedia Delivery

This paper proposes a novel approach, based on unequal error protection, to enhance rateless codes with progressive recovery for layered multimedia delivery. With a parallel encoding structure, the proposed Progressive Rateless codes (PRC) assign unequal redundancy to each layer in accordance with their importance. Each output symbol contains information from all layers, and thus the stream layers can be recovered progressively at the expected received ratios of output symbols. Furthermore, the dependency between layers is naturally considered. The performance of the PRC is evaluated and compared with some related UEP approaches. Results show that our PRC approach provides better recovery performance with lower overhead both theoretically and numerically.

cs.IT↗

arXiv2012

Unequal Error Protected JPEG 2000 Broadcast Scheme with Progressive Fountain Codes

This paper proposes a novel scheme, based on progressive fountain codes, for broadcasting JPEG 2000 multimedia. In such a broadcast scheme, progressive resolution levels of images/video have been unequally protected when transmitted using the proposed progressive fountain codes. With progressive fountain codes applied in the broadcast scheme, the resolutions of images (JPEG 2000) or videos (MJPEG 2000) received by different users can be automatically adaptive to their channel qualities, i.e. the users with good channel qualities are possible to receive the high resolution images/vedio while the users with bad channel qualities may receive low resolution images/vedio. Finally, the performance of the proposed scheme is evaluated with the MJPEG 2000 broadcast prototype.

cs.IT↗