Pub Date : 2024-02-22DOI: 10.1109/TBC.2024.3358756
Li Yang;Jianzhang Liu;Shufeng Li;Deyou Zhang;Zhiping Xia
Video streaming services have gradually become the dominant use case on the Internet, where people’s focus has shifted from a service-centric approach to a user-centric approach. This paper proposes a continuous Quality of Experience (QoE) evaluation method based on Quality of Service (QoS) parameters, which combines the advantages of subjective and objective research methods. The proposed method can accurately calculate the achievable QoE based on user QoS. Additionally, we develop a QoE-Ensemble MLP prediction model, employing ensemble learning and MLP techniques, to overcome limitations of the QoE evaluation method and accurately predict user QoE based on QoS parameters. Furthermore, we propose a low-complexity network bandwidth allocation algorithm based on a QoE prediction model to help service providers minimize network bandwidth waste while meeting user QoE requirements. Finally, the experiments show that our QoE evaluation model and network bandwidth allocation algorithm have better performance. And according to the result analysis, we also got the connection between QoS parameters and QoE.
{"title":"Enhancing User Experience in Ultra HD Cloud Performing Arts Live Streaming: A QoS-to-QoE Mapping Approach","authors":"Li Yang;Jianzhang Liu;Shufeng Li;Deyou Zhang;Zhiping Xia","doi":"10.1109/TBC.2024.3358756","DOIUrl":"10.1109/TBC.2024.3358756","url":null,"abstract":"Video streaming services have gradually become the dominant use case on the Internet, where people’s focus has shifted from a service-centric approach to a user-centric approach. This paper proposes a continuous Quality of Experience (QoE) evaluation method based on Quality of Service (QoS) parameters, which combines the advantages of subjective and objective research methods. The proposed method can accurately calculate the achievable QoE based on user QoS. Additionally, we develop a QoE-Ensemble MLP prediction model, employing ensemble learning and MLP techniques, to overcome limitations of the QoE evaluation method and accurately predict user QoE based on QoS parameters. Furthermore, we propose a low-complexity network bandwidth allocation algorithm based on a QoE prediction model to help service providers minimize network bandwidth waste while meeting user QoE requirements. Finally, the experiments show that our QoE evaluation model and network bandwidth allocation algorithm have better performance. And according to the result analysis, we also got the connection between QoS parameters and QoE.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"413-428"},"PeriodicalIF":4.5,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Panoramic video undergoes projection onto a two-dimensional plane for compression and subsequent back-projection onto a sphere for display. This process introduces inconsistency between compression distortion and perceived spherical distortion, which causes a serious loss in coding efficiency. Meanwhile, the existing independent rate-distortion optimization (RDO) model for spherical distortion solely accounts for the current coding frame and neglects its influence on subsequent frames, which leads to sub-optimal coding performance. To this end, we propose a spherical distortion temporal propagation and spatial mapping model for efficient panoramic video coding. First, a zero-delay spherical distortion backward propagation chain is established in the temporal domain, and distortion impact factors are computed. Then, an accurate spatial mapping relationship between spherical distortion and coding distortion is constructed, along with the calculation of spatial mapping weights. Finally, these components are integrated into spherical RDO. The experimental results demonstrated the effectiveness of the proposed algorithm. Compared to the versatile video coding test model (VTM-14.0) with a 360Lib extension under low-delay P frame and B frame configurations, the proposed algorithm achieves bitrate savings of 9.4% (up to 19.4%) and 8.5% (up to 19.0%) by using WSPSNR as the distortion evaluation index, respectively. Additionally, the coding time was reduced by 14.53% and 15.65%, respectively.
全景视频先投影到二维平面上进行压缩，然后再反投影到球面上进行显示。这一过程会导致压缩失真与感知球面失真不一致，从而严重降低编码效率。同时，现有的球形失真独立速率-失真优化（RDO）模型只考虑当前编码帧，忽略了其对后续帧的影响，导致编码性能未达到最佳。为此，我们提出了一种球形失真时间传播和空间映射模型，用于高效的全景视频编码。首先，在时域建立零延迟球形失真后向传播链，并计算失真影响因子。然后，构建球形失真与编码失真之间的精确空间映射关系，并计算空间映射权重。最后，将这些组件集成到球形 RDO 中。实验结果证明了所提算法的有效性。在低延迟 P 帧和 B 帧配置下，与带有 360Lib 扩展的通用视频编码测试模型（VTM-14.0）相比，以 WSPSNR 作为失真评估指标，所提算法分别节省了 9.4% （最高 19.4%）和 8.5%（最高 19.0%）的比特率。此外，编码时间也分别缩短了 14.53% 和 15.65%。
{"title":"Spherical Distortion Temporal Propagation and Spatial Mapping Model for Efficient Panoramic Video Coding","authors":"Xu Yang;Minfeng Huang;Hongwei Guo;Shengxi Li;Lei Luo;Ce Zhu","doi":"10.1109/TBC.2024.3358749","DOIUrl":"10.1109/TBC.2024.3358749","url":null,"abstract":"Panoramic video undergoes projection onto a two-dimensional plane for compression and subsequent back-projection onto a sphere for display. This process introduces inconsistency between compression distortion and perceived spherical distortion, which causes a serious loss in coding efficiency. Meanwhile, the existing independent rate-distortion optimization (RDO) model for spherical distortion solely accounts for the current coding frame and neglects its influence on subsequent frames, which leads to sub-optimal coding performance. To this end, we propose a spherical distortion temporal propagation and spatial mapping model for efficient panoramic video coding. First, a zero-delay spherical distortion backward propagation chain is established in the temporal domain, and distortion impact factors are computed. Then, an accurate spatial mapping relationship between spherical distortion and coding distortion is constructed, along with the calculation of spatial mapping weights. Finally, these components are integrated into spherical RDO. The experimental results demonstrated the effectiveness of the proposed algorithm. Compared to the versatile video coding test model (VTM-14.0) with a 360Lib extension under low-delay P frame and B frame configurations, the proposed algorithm achieves bitrate savings of 9.4% (up to 19.4%) and 8.5% (up to 19.0%) by using WSPSNR as the distortion evaluation index, respectively. Additionally, the coding time was reduced by 14.53% and 15.65%, respectively.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"654-666"},"PeriodicalIF":4.5,"publicationDate":"2024-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947558","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
As the latest video coding standard, Versatile Video Coding (VVC) is highly efficient at the cost of very high coding complexity, which seriously hinders its practical application. Therefore, it is very crucial to improve its coding speed. In this paper, we propose a learning-based fast split mode (SM) and directional mode (DM) decision algorithm for VVC intra prediction using a deep learning approach. Specifically, given the observation that the SM distributions of coding units (CUs) of different sizes are significantly distinct, we first design the neural networks separately and train the SM models for all CUs of different sizes to obtain the probability of SMs and skip the unlikely ones. Second, given a similar observation that the DM distributions of CUs of different sizes are distinct, we design neural networks to train the DM models for all CUs of different sizes separately to obtain the probabilities of DMs, and then adaptively select candidate DMs based on probabilities of their located SMs. Third, after an SM is checked, we select its probability, residual coefficients, rate-distortion (RD) cost, etc. as features, and design a lightweight neural network (LNN) model to early terminate SM selection. Experimental results demonstrate that the proposed algorithm can reduce the encoding time of VVC by 70.73% with 2.44% increase in Bjøntegaard delta bit-rate (BDBR) on average.
作为最新的视频编码标准，多功能视频编码（VVC）具有极高的编码效率，但代价是极高的编码复杂度，这严重阻碍了它的实际应用。因此，提高其编码速度至关重要。在本文中，我们提出了一种基于学习的快速分割模式（SM）和定向模式（DM）决策算法，利用深度学习方法进行 VVC 内部预测。具体来说，鉴于观察到不同大小的编码单元（CU）的 SM 分布明显不同，我们首先分别设计神经网络，并对所有不同大小的 CU 训练 SM 模型，以获得 SM 的概率并跳过不可能的 SM。其次，鉴于不同大小的 CU 的 DM 分布具有类似的观察结果，我们设计神经网络，分别训练所有不同大小的 CU 的 DM 模型，以获得 DM 的概率，然后根据其所在 SM 的概率自适应地选择候选 DM。第三，在检查出 SM 后，我们选择其概率、残差系数、速率失真（RD）成本等作为特征，并设计轻量级神经网络（LNN）模型来提前终止 SM 选择。实验结果表明，所提出的算法可将 VVC 的编码时间缩短 70.73%，而 Bjøntegaard delta 比特率（BDBR）平均提高 2.44%。
{"title":"Learning-Based Fast Splitting and Directional Mode Decision for VVC Intra Prediction","authors":"Yuanyuan Huang;Junyi Yu;Dayong Wang;Xin Lu;Frederic Dufaux;Hui Guo;Ce Zhu","doi":"10.1109/TBC.2024.3360729","DOIUrl":"10.1109/TBC.2024.3360729","url":null,"abstract":"As the latest video coding standard, Versatile Video Coding (VVC) is highly efficient at the cost of very high coding complexity, which seriously hinders its practical application. Therefore, it is very crucial to improve its coding speed. In this paper, we propose a learning-based fast split mode (SM) and directional mode (DM) decision algorithm for VVC intra prediction using a deep learning approach. Specifically, given the observation that the SM distributions of coding units (CUs) of different sizes are significantly distinct, we first design the neural networks separately and train the SM models for all CUs of different sizes to obtain the probability of SMs and skip the unlikely ones. Second, given a similar observation that the DM distributions of CUs of different sizes are distinct, we design neural networks to train the DM models for all CUs of different sizes separately to obtain the probabilities of DMs, and then adaptively select candidate DMs based on probabilities of their located SMs. Third, after an SM is checked, we select its probability, residual coefficients, rate-distortion (RD) cost, etc. as features, and design a lightweight neural network (LNN) model to early terminate SM selection. Experimental results demonstrate that the proposed algorithm can reduce the encoding time of VVC by 70.73% with 2.44% increase in Bjøntegaard delta bit-rate (BDBR) on average.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"681-692"},"PeriodicalIF":4.5,"publicationDate":"2024-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947424","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Timing mismatch among signals of different receiving antennas degrades the maximum ratio combining (MRC) performance in broadcasting systems, and thus limits the output signal-to-noise ratio (SNR) after MRC. To circumvent this limitation, in this paper, a time-robust MRC (TR-MRC) receiver for broadcasting reception enhancement is designed by employing the tapped-delay-line (TDL) structure, where the weight matrix is derived as iterated-form and the average output SNR is calculated as closed-form. Simulation results confirm that the proposed TR-MRC receiver can improve the average output SNR and reduce the bit error rate (BER) under imperfect time alignment, especially for large number of receiving antennas and high input SNR. However, blindly increasing the TDL order can not bring significant improvement of the TR-MRC performance with unknown timing mismatch, showing a trade-off between the output SNR and the implementation complexity introduced by increasing TDL order.
{"title":"Time-Robust MRC Design for Broadcasting Reception Enhancement","authors":"Wenbo Guo;Hongzhi Zhao;Jiaxin Du;Shihai Shao;Youxi Tang","doi":"10.1109/TBC.2024.3353569","DOIUrl":"10.1109/TBC.2024.3353569","url":null,"abstract":"Timing mismatch among signals of different receiving antennas degrades the maximum ratio combining (MRC) performance in broadcasting systems, and thus limits the output signal-to-noise ratio (SNR) after MRC. To circumvent this limitation, in this paper, a time-robust MRC (TR-MRC) receiver for broadcasting reception enhancement is designed by employing the tapped-delay-line (TDL) structure, where the weight matrix is derived as iterated-form and the average output SNR is calculated as closed-form. Simulation results confirm that the proposed TR-MRC receiver can improve the average output SNR and reduce the bit error rate (BER) under imperfect time alignment, especially for large number of receiving antennas and high input SNR. However, blindly increasing the TDL order can not bring significant improvement of the TR-MRC performance with unknown timing mismatch, showing a trade-off between the output SNR and the implementation complexity introduced by increasing TDL order.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"747-752"},"PeriodicalIF":4.5,"publicationDate":"2024-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947622","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Recent methods that project images into deep feature spaces to evaluate quality degradation have produced inefficient results due to biased mappings; i.e., these projections are not aligned with the perceptions of humans. In this paper, we develop a hyperdebiasing framework to address such bias in full-reference image quality assessment. First, we perform orthogonal Tucker decomposition on the top of feature tensors extracted by a feature extraction network to project features into a robust content-agnostic space and effectively eliminate the bias caused by subtle image perturbations. Second, we propose a hypernetwork in which the content-aware parameters are produced for reprojecting features in a deep subspace for quality prediction. By leveraging the content diversity of large-scale blind-reference datasets, the perception rule between image content and image quality is established. Third, a quality prediction network is proposed by combining debiased content-aware and content-agnostic features to predict the final image quality score. To demonstrate the efficacy of our proposed method, we conducted numerous experiments on comprehensive databases. The experimental results validate that our method achieves state-of-the-art performance in predicting image quality.
{"title":"HDIQA: A Hyper Debiasing Framework for Full Reference Image Quality Assessment","authors":"Mingliang Zhou;Heqiang Wang;Xuekai Wei;Yong Feng;Jun Luo;Huayan Pu;Jinglei Zhao;Liming Wang;Zhigang Chu;Xin Wang;Bin Fang;Zhaowei Shang","doi":"10.1109/TBC.2024.3353573","DOIUrl":"10.1109/TBC.2024.3353573","url":null,"abstract":"Recent methods that project images into deep feature spaces to evaluate quality degradation have produced inefficient results due to biased mappings; i.e., these projections are not aligned with the perceptions of humans. In this paper, we develop a hyperdebiasing framework to address such bias in full-reference image quality assessment. First, we perform orthogonal Tucker decomposition on the top of feature tensors extracted by a feature extraction network to project features into a robust content-agnostic space and effectively eliminate the bias caused by subtle image perturbations. Second, we propose a hypernetwork in which the content-aware parameters are produced for reprojecting features in a deep subspace for quality prediction. By leveraging the content diversity of large-scale blind-reference datasets, the perception rule between image content and image quality is established. Third, a quality prediction network is proposed by combining debiased content-aware and content-agnostic features to predict the final image quality score. To demonstrate the efficacy of our proposed method, we conducted numerous experiments on comprehensive databases. The experimental results validate that our method achieves state-of-the-art performance in predicting image quality.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"545-554"},"PeriodicalIF":4.5,"publicationDate":"2024-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Elevating traditional 2-dimensional (2D) plane display to 4-dimensional (4D) light field display can significantly enhance users’ immersion and realism, because light field image (LFI) provides various visual cues in terms of multi-view disparity, motion disparity, and selective focus. Therefore, it is crucial to establish a light field image quality assessment (LF-IQA) model that aligns with human visual perception characteristics. However, it has always been a challenge to evaluate the perceptual quality of multiple light field visual cues simultaneously and consistently. To this end, this paper proposes a Transformer-based explicit learning of light field geometry for the no-reference light field image quality assessment. Specifically, to explicitly learn the light field epipolar geometry, we stack up light field sub-aperture images (SAIs) to form four SAI stacks according to four specific light field angular directions, and use a sub-grouping strategy to hierarchically learn the local and global light field geometric features. Then, a Transformer encoder with a spatial-shift tokenization strategy is applied to learn structure-aware light field geometric distortion representation, which is used to regress the final quality score. Evaluation experiments are carried out on three commonly used light field image quality assessment datasets: Win5-LID, NBU-LF1.0, and MPI-LFA. Experimental results demonstrate that our model outperforms state-of-the-art methods and exhibits a high correlation with human perception. The source code is publicly available at https://github.com/windyz77/GeoNRLFIQA .
将传统的二维（2D）平面显示提升到四维（4D）光场显示，可以显著增强用户的沉浸感和真实感，因为光场图像（LFI）提供了多视角差异、运动差异和选择性聚焦等多种视觉线索。因此，建立一个符合人类视觉感知特征的光场图像质量评估（LF-IQA）模型至关重要。然而，如何同时、一致地评估多个光场视觉线索的感知质量一直是个难题。为此，本文提出了一种基于变换器的光场几何显式学习方法，用于无参照光场图像质量评估。具体来说，为了显式学习光场外极几何，我们将光场子孔径图像（SAI）按照四个特定的光场角度方向堆叠成四个 SAI 堆栈，并使用子分组策略分层学习局部和全局光场几何特征。然后，采用空间偏移标记化策略的变换器编码器学习结构感知光场几何失真表示，并以此回归最终质量得分。评估实验在三个常用的光场图像质量评估数据集上进行：Win5-LID、NBU-LF1.0 和 MPI-LFA。实验结果表明，我们的模型优于最先进的方法，并且与人类感知具有很高的相关性。源代码可通过 https://github.com/windyz77/GeoNRLFIQA 公开获取。
{"title":"Transformer-Based Light Field Geometry Learning for No-Reference Light Field Image Quality Assessment","authors":"Lili Lin;Siyu Bai;Mengjia Qu;Xuehui Wei;Luyao Wang;Feifan Wu;Biao Liu;Wenhui Zhou;Ercan Engin Kuruoglu","doi":"10.1109/TBC.2024.3353579","DOIUrl":"10.1109/TBC.2024.3353579","url":null,"abstract":"Elevating traditional 2-dimensional (2D) plane display to 4-dimensional (4D) light field display can significantly enhance users’ immersion and realism, because light field image (LFI) provides various visual cues in terms of multi-view disparity, motion disparity, and selective focus. Therefore, it is crucial to establish a light field image quality assessment (LF-IQA) model that aligns with human visual perception characteristics. However, it has always been a challenge to evaluate the perceptual quality of multiple light field visual cues simultaneously and consistently. To this end, this paper proposes a Transformer-based explicit learning of light field geometry for the no-reference light field image quality assessment. Specifically, to explicitly learn the light field epipolar geometry, we stack up light field sub-aperture images (SAIs) to form four SAI stacks according to four specific light field angular directions, and use a sub-grouping strategy to hierarchically learn the local and global light field geometric features. Then, a Transformer encoder with a spatial-shift tokenization strategy is applied to learn structure-aware light field geometric distortion representation, which is used to regress the final quality score. Evaluation experiments are carried out on three commonly used light field image quality assessment datasets: Win5-LID, NBU-LF1.0, and MPI-LFA. Experimental results demonstrate that our model outperforms state-of-the-art methods and exhibits a high correlation with human perception. The source code is publicly available at \u0000<uri>https://github.com/windyz77/GeoNRLFIQA</uri>\u0000.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"597-606"},"PeriodicalIF":4.5,"publicationDate":"2024-01-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Inter-tower communications networks (ITCN) and wireless in-band distribution links (IDL) reuse the same broadcast spectrum for establishing communications links between transmitter towers and for wireless backhaul by multiplexing the ITCN/IDL signals with the broadcast signal into a frame for transmission. In single-frequency networks (SFN) environment, where all the transmitter towers broadcast the same preamble and in-band pilots for improving TV coverage and received signal strength, receiving the desired ITCN/IDL signal from a specific transmitter is challenging with the conventional channel estimation techniques. In the Advanced Television Standard Committee (ATSC) 3.0, one unique transmitter identification (TxID) sequence, a spread sequence overlaid with the preamble signal, is assigned for each transmitter for the purpose of SFN planning and synchronization. By using the TxID sequence, the channel estimation of a specific transmitter becomes feasible. However, the accuracy of existing TxID-based channel estimation is limited due to interferences from the preamble signal and the co-channel TxIDs, as well as the non-orthogonality of the TxID sequence. Several high-accuracy channel estimation schemes based on the TxID sequence are proposed in this paper, which enable IDL and ITCN with very high data rate transmission, e.g., 1024 QAM modulation.
{"title":"High Accuracy Channel Estimation With TxID Sequence in ATSC 3.0 SFN","authors":"Zhihong Hunter Hong;Yiyan Wu;Wei Li;Liang Zhang;Zhiwen Zhu;Sung-Ik Park;Namho Hur;Eneko Iradier;Jon Montalban","doi":"10.1109/TBC.2024.3353577","DOIUrl":"10.1109/TBC.2024.3353577","url":null,"abstract":"Inter-tower communications networks (ITCN) and wireless in-band distribution links (IDL) reuse the same broadcast spectrum for establishing communications links between transmitter towers and for wireless backhaul by multiplexing the ITCN/IDL signals with the broadcast signal into a frame for transmission. In single-frequency networks (SFN) environment, where all the transmitter towers broadcast the same preamble and in-band pilots for improving TV coverage and received signal strength, receiving the desired ITCN/IDL signal from a specific transmitter is challenging with the conventional channel estimation techniques. In the Advanced Television Standard Committee (ATSC) 3.0, one unique transmitter identification (TxID) sequence, a spread sequence overlaid with the preamble signal, is assigned for each transmitter for the purpose of SFN planning and synchronization. By using the TxID sequence, the channel estimation of a specific transmitter becomes feasible. However, the accuracy of existing TxID-based channel estimation is limited due to interferences from the preamble signal and the co-channel TxIDs, as well as the non-orthogonality of the TxID sequence. Several high-accuracy channel estimation schemes based on the TxID sequence are proposed in this paper, which enable IDL and ITCN with very high data rate transmission, e.g., 1024 QAM modulation.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"391-400"},"PeriodicalIF":4.5,"publicationDate":"2024-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-30DOI: 10.1109/TBC.2024.3353568
Linyao Gao;Zhu Li;Lizhi Hou;Yiling Xu;Jun Sun
Video-based point cloud compression (V-PCC) has achieved remarkable compression efficiency, which converts point clouds into videos and leverages video codecs for coding. For lossy compression, the undesirable artifacts of attribute images always degrade the point clouds attribute reconstruction quality. In this paper, we propose an Occupancy-assisted Compression Artifact Removal Network (OCARNet) to remove the distortions of V-PCC decoded attribute images for high-quality point cloud attribute reconstruction. Specifically, the occupancy information is fed into network as a prior knowledge to provide more spatial and structural information and to assist in eliminating the distortions of the texture regions. To aggregate the occupancy information effectively, we design a multi-level feature fusion framework with Channel-Spatial Attention based Residual Blocks (CSARB), where the short and long residual connections are jointly employed to capture the local context and long-range dependency. Besides, we propose a Masked Mean Square Error (MMSE) loss function based on the occupancy information to train our proposed network to focus on estimating the attribute artifacts of the occupied regions. To the best of our knowledge, this is the first learning-based attribute artifact removal method for V-PCC. Experimental results demonstrate that our framework outperforms existing state-of-the-art methods and shows the effectiveness on both objective and subjective quality comparisons.
{"title":"Occupancy-Assisted Attribute Artifact Reduction for Video-Based Point Cloud Compression","authors":"Linyao Gao;Zhu Li;Lizhi Hou;Yiling Xu;Jun Sun","doi":"10.1109/TBC.2024.3353568","DOIUrl":"10.1109/TBC.2024.3353568","url":null,"abstract":"Video-based point cloud compression (V-PCC) has achieved remarkable compression efficiency, which converts point clouds into videos and leverages video codecs for coding. For lossy compression, the undesirable artifacts of attribute images always degrade the point clouds attribute reconstruction quality. In this paper, we propose an Occupancy-assisted Compression Artifact Removal Network (OCARNet) to remove the distortions of V-PCC decoded attribute images for high-quality point cloud attribute reconstruction. Specifically, the occupancy information is fed into network as a prior knowledge to provide more spatial and structural information and to assist in eliminating the distortions of the texture regions. To aggregate the occupancy information effectively, we design a multi-level feature fusion framework with Channel-Spatial Attention based Residual Blocks (CSARB), where the short and long residual connections are jointly employed to capture the local context and long-range dependency. Besides, we propose a Masked Mean Square Error (MMSE) loss function based on the occupancy information to train our proposed network to focus on estimating the attribute artifacts of the occupied regions. To the best of our knowledge, this is the first learning-based attribute artifact removal method for V-PCC. Experimental results demonstrate that our framework outperforms existing state-of-the-art methods and shows the effectiveness on both objective and subjective quality comparisons.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"667-680"},"PeriodicalIF":4.5,"publicationDate":"2024-01-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947557","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-25DOI: 10.1109/TBC.2024.3349768
Sujie Shao;Linlin Zhang;Fei Qi
With the improvement of terminal intelligence and the enrichment of digital content, terminal density is showing an explosive growth trend, and the traffic carried by IPTV and other services is rapidly increasing. HDHB WLAN (High-Density High-Bandwidth Wireless LAN) is becoming a dominant form of wireless LAN. However, the RSSI-based access mode has led to a notable load imbalance, and the resource competition mode based on random access intensifies the difficulty of access resource acquisition, which exacerbates the traffic challenges faced by WLAN. IEEE 802.11ax somewhat alleviates traffic pressure, but it does not fundamentally solve these problems. This paper introduces an access optimization mechanism for the 802.11ax HDHB WLAN, which aims to achieve load balancing while considering competition avoidance, alleviating the pressure of IPTV traffic. First, an 802.11ax access optimization architecture for HDHB WLAN is constructed, aimed at alleviating traffic pressure and meeting the quality requirements of IPTV and other services by modifying access processes of terminals. Next, a terminal information acquisition and interactive access control strategy based on the trigger frame is devised to obtain accurate parameter information and facilitate orderly concurrent access control for high-density terminals. Additionally, a load balancing and competition avoidance oriented access control method for HDHB WLAN is proposed, including an access optimization control model, and an access strategy generation algorithm based on the Improved DQN algorithm. Finally, simulation results show that the global throughput and load balancing of HDHB WLAN are improved, consequently reducing overall WLAN traffic pressure.
{"title":"Access Optimization in 802.11ax WLAN for Load Balancing and Competition Avoidance of IPTV Traffic","authors":"Sujie Shao;Linlin Zhang;Fei Qi","doi":"10.1109/TBC.2024.3349768","DOIUrl":"10.1109/TBC.2024.3349768","url":null,"abstract":"With the improvement of terminal intelligence and the enrichment of digital content, terminal density is showing an explosive growth trend, and the traffic carried by IPTV and other services is rapidly increasing. HDHB WLAN (High-Density High-Bandwidth Wireless LAN) is becoming a dominant form of wireless LAN. However, the RSSI-based access mode has led to a notable load imbalance, and the resource competition mode based on random access intensifies the difficulty of access resource acquisition, which exacerbates the traffic challenges faced by WLAN. IEEE 802.11ax somewhat alleviates traffic pressure, but it does not fundamentally solve these problems. This paper introduces an access optimization mechanism for the 802.11ax HDHB WLAN, which aims to achieve load balancing while considering competition avoidance, alleviating the pressure of IPTV traffic. First, an 802.11ax access optimization architecture for HDHB WLAN is constructed, aimed at alleviating traffic pressure and meeting the quality requirements of IPTV and other services by modifying access processes of terminals. Next, a terminal information acquisition and interactive access control strategy based on the trigger frame is devised to obtain accurate parameter information and facilitate orderly concurrent access control for high-density terminals. Additionally, a load balancing and competition avoidance oriented access control method for HDHB WLAN is proposed, including an access optimization control model, and an access strategy generation algorithm based on the Improved DQN algorithm. Finally, simulation results show that the global throughput and load balancing of HDHB WLAN are improved, consequently reducing overall WLAN traffic pressure.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"401-412"},"PeriodicalIF":4.5,"publicationDate":"2024-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947562","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-01-25DOI: 10.1109/TBC.2024.3349773
Alireza Esmaeilzehi;Yang Ou;M. Omair Ahmad;M. N. S. Swamy
Enhancing the quality of the images acquired under the water environments is crucial in many broadcast technologies. As the richness of the features generated by deep underwater image enhancement networks improves, the visual signals with higher qualities can be yielded. In view of this, in this paper, we propose a new deep network for the task of underwater image enhancement, in which the network feature generation process is guided by the prior information obtained from various underwater medium transmission map and atmospheric light estimation methods. Further, in order to obtain high values for different image quality assessment metrics associated with the images produced by the proposed network, we introduce a multi-stage training process for our network. In the first stage, the proposed network is trained with the conventional supervised learning technique, whereas, in the second stage, the training process of the network is carried out by the adversarial learning technique. Finally, in the third stage, the training of the network obtained by the conventional supervised learning is continued by the guidance of the one trained by the adversarial learning technique. In the development of the adversarial learning-based stage of our network, we propose a novel multi-discriminator generative adversarial network, which is able to produce images with more realistic textures and structures. The proposed multi-discriminator generative adversarial network employs the discrimination process between the real and fake data in various underwater environment color spaces. The results of different experimentations show the effectiveness of the proposed scheme in restoring the high-quality images compared to the other state-of-the-art deep underwater image enhancement networks.
{"title":"DMML: Deep Multi-Prior and Multi-Discriminator Learning for Underwater Image Enhancement","authors":"Alireza Esmaeilzehi;Yang Ou;M. Omair Ahmad;M. N. S. Swamy","doi":"10.1109/TBC.2024.3349773","DOIUrl":"10.1109/TBC.2024.3349773","url":null,"abstract":"Enhancing the quality of the images acquired under the water environments is crucial in many broadcast technologies. As the richness of the features generated by deep underwater image enhancement networks improves, the visual signals with higher qualities can be yielded. In view of this, in this paper, we propose a new deep network for the task of underwater image enhancement, in which the network feature generation process is guided by the prior information obtained from various underwater medium transmission map and atmospheric light estimation methods. Further, in order to obtain high values for different image quality assessment metrics associated with the images produced by the proposed network, we introduce a multi-stage training process for our network. In the first stage, the proposed network is trained with the conventional supervised learning technique, whereas, in the second stage, the training process of the network is carried out by the adversarial learning technique. Finally, in the third stage, the training of the network obtained by the conventional supervised learning is continued by the guidance of the one trained by the adversarial learning technique. In the development of the adversarial learning-based stage of our network, we propose a novel multi-discriminator generative adversarial network, which is able to produce images with more realistic textures and structures. The proposed multi-discriminator generative adversarial network employs the discrimination process between the real and fake data in various underwater environment color spaces. The results of different experimentations show the effectiveness of the proposed scheme in restoring the high-quality images compared to the other state-of-the-art deep underwater image enhancement networks.","PeriodicalId":13159,"journal":{"name":"IEEE Transactions on Broadcasting","volume":"70 2","pages":"637-653"},"PeriodicalIF":4.5,"publicationDate":"2024-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139947564","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：