This work investigates video encoding parallelization techniques based on the VVC standard, using the open and optimized encoder software implementation VVenC. Modern multi-processor systems offer significant opportunities for accelerating video encoding. By employing a proposed combination of parallelization methods, the VVenC encoder achieves an acceleration factor of up to 22 compared to single-threaded mode on a 32-core system, with potential increases to 27x at higher bitrates. Building upon prior work on inter-frame parallelization (IFP), the study introduces frame region-based synchronization, enabling further acceleration of up to 10%. Beyond that, the study demonstrates extending frame parallelization beyond Group of Pictures (GOP) boundaries, which improves IFP speed up by 37% and 11% at HD and UHD resolutions, respectively. Additional combinations with other VVC parallelization tools, such as tiles and VVC Wavefront Parallel Processing (WPP), are also explored. The paper provides a comprehensive analysis of parallelization challenges and highlights areas for further improvement.

Existing 3D Gaussian Splatting simplification methods commonly use importance scores, such as blending weights or sensitivity, to identify redundant Gaussians. However, these scores are not driven by visual error metrics, often leading to suboptimal trade-offs between compactness and rendering fidelity. We present GaussianPOP, a principled simplification framework based on analytical Gaussian error quantification. Our key contribution is a novel error criterion, derived directly from the 3DGS rendering equation, that precisely measures each Gaussian's contribution to the rendered image. By introducing a highly efficient algorithm, our framework enables practical error calculation in a single forward pass. The framework is both accurate and flexible, supporting on-training pruning as well as post-training simplification via iterative error re-quantification for improved stability. Experimental results show that our method consistently outperforms existing state-of-the-art pruning methods across both application scenarios, achieving a superior trade-off between model compactness and high rendering quality.

3D Gaussian Splatting (3DGS) has revolutionized real-time rendering capabilities, yet its substantial storage footprint significantly hinders deployment on resource-constrained mobile devices. While implicit representations like CodecGS effectively reduce storage by projecting Gaussian attributes onto feature planes, this framework relies on 16-bit data formats, rendering it incompatible with standard mobile hardware video decoders. This paper presents a practical, low-latency decoding pipeline for CodecGS, specifically optimized for the Meta Quest 3 within the Unity engine. We bridge the compatibility gap by integrating Unity's native FFmpeg for hardware-accelerated decoding and the Unity Sentis engine for on-device MLP inference. Crucially, we propose a strategic shift to 8-bit quantization for strict compatibility with mobile hardware constraints. By leveraging this pipeline, this paper demonstrates the end-to-end workflow for streaming and rendering CodecGS-based content on standalone VR headsets.

Coding of Gaussian splats has drawn the attention of academia and standardization bodies lately. A commonly used approach involves projecting explicit 3D attributes onto 2D planes of a video and compressing it with existing video coding standards. However, such an approach produces excessively high sample rates, resulting in a significant bottleneck that might make it unsuitable for existing hardware decoders if the number of splats is high. This paper presents a framework that utilizes an implicit representation, where Gaussian splat attributes are represented by compact feature planes. By reducing the dependency of video resolution on the number of splats, our approach significantly reduces the required sample rate, making it suitable for deployed devices. In addition, the proposed framework achieves a superior rate-distortion trade-off, providing high-fidelity reconstruction at low bitrates without the excessive sample rates associated with the conventional video-based anchor.

3D Gaussian Splatting is a recognized method for 3D scene representation, known for its high rendering quality and speed. However, its substantial data requirements present challenges for practical applications. In this paper, we introduce an efficient compression technique that significantly reduces storage overhead by using compact representation. We propose a unified architecture that combines point cloud data and feature planes through a progressive tri-plane structure. Our method utilizes 2D feature planes, enabling continuous spatial representation. To further optimize these representations, we incorporate entropy modeling in the frequency domain, specifically designed for standard video codecs. We also propose channel-wise bit allocation to achieve a better trade-off between bitrate consumption and feature plane representation. Consequently, our model effectively leverages spatial correlations within the feature planes to enhance rate-distortion performance using standard, non-differentiable video codecs. Experimental results demonstrate that our method outperforms existing methods in data compactness while maintaining high rendering quality. Our project page is available at https://fraunhoferhhi.github.io/CodecGS

Vehicle-to-everything (V2X) communication is a key technology for enabling intelligent transportation systems (ITS) that can improve road safety, traffic efficiency, and environmental sustainability. Among the various V2X applications, platooning is one of the most promising ones, as it allows a group of vehicles to travel closely together at high speeds, reducing fuel consumption and emissions. However, it poses significant challenges for wireless communication, such as high reliability and low latency. In this paper, we evaluate the benefits of group scheduling, also referred to as Mode 2d, which is based on a distributed and scheduled resource allocation scheme that allows the group of cars to select resources from a configured pool without network assistance. We evaluated the scheme through simulations, and the results show that this approach can meet the reliability, low latency, and data rate requirements for platooning.

Explicit feature-grid based NeRF models have shown promising results in terms of rendering quality and significant speed-up in training. However, these methods often require a significant amount of data to represent a single scene or object. In this work, we present a compression model that aims to minimize the entropy in the frequency domain in order to effectively reduce the data size. First, we propose using the discrete cosine transform (DCT) on the tensorial radiance fields to compress the feature-grid. This feature-grid is transformed into coefficients, which are then quantized and entropy encoded, following a similar approach to the traditional video coding pipeline. Furthermore, to achieve a higher level of sparsity, we propose using an entropy parameterization technique for the frequency domain, specifically for DCT coefficients of the feature-grid. Since the transformed coefficients are optimized during the training phase, the proposed model does not require any fine-tuning or additional information. Our model only requires a lightweight compression pipeline for encoding and decoding, making it easier to apply volumetric radiance field methods for real-world applications. Experimental results demonstrate that our proposed frequency domain entropy model can achieve superior compression performance across various datasets. The source code will be made publicly available.

The Versatile Video Coding (VVC) standard promises high compression efficiency for diverse content types. Based on VVenC, an open and optimized VVC software video encoder, this work presents an inter-frame parallelization (IFP) method designed to exploit the processing power of modern platforms featuring a high number of computing cores. Encoding an ultrahigh definition video on a 32-core machine with the VVenC's faster preset, the proposed method shows more than 20% increase in encoder speed while only a 1% decrease in compression efficiency compared to the default multi-threading mode. In comparison to single-threaded mode, it corresponds to a speedup factor of 18, up from 15x achievable with the previous parallelization scheme. Furthermore, the synergy of the developed inter-frame parallelization technique with other parallelization methods is explored, including tiles and VVC wavefront parallel processing (WPP). The combination of these approaches enables a notable speedup factor of 21, albeit with a trade-off in coding efficiency. With a focus on VVC, this research contributes to the ongoing discourse on video coding optimization, providing valuable insights into possible pitfalls and the potential gains achievable through efficient parallelization techniques on high-core platforms.

In this work, we propose novel HARQ prediction schemes for Cloud RANs (C-RANs) that use feedback over a rate-limited feedback channel (2 - 6 bits) from the Remote Radio Heads (RRHs) to predict at the User Equipment (UE) the decoding outcome at the BaseBand Unit (BBU) ahead of actual decoding. In particular, we propose a Dual Autoencoding 2-Stage Gaussian Mixture Model (DA2SGMM) that is trained in an end-to-end fashion over the whole C-RAN setup. Using realistic link-level simulations in the sub-THz band at 100 GHz, we show that the novel DA2SGMM HARQ prediction scheme clearly outperforms all other adapted and state-of-the-art schemes. The DA2SGMM shows a superior performance in terms of blockage detection as well as HARQ prediction in the no-blockage and single-blockage cases. In particular, the DA2SGMM with 4~bit feedback achieves a more than 200 % higher throughput in average compared to its best alternative. Compared to regular HARQ, the DA2SGMM reduces the maximum transmission latency by more than 72.4 %, while maintaining more than 75 % of the throughput in the no-blockage scenario. In the single-blockage scenario, DA2SGMM significantly increases the throughput for most of the evaluated Signal-to-Noise-Ratios (SNRs) compared to regular HARQ.

Film grain is a distinctive visual characteristic cherished by filmmakers and cinephiles for its ability to evoke nostalgia and artistic aesthetics. However, faithful preservation of film grain during encoding poses unique challenges. Film grain introduces random noise, complicating traditional compression techniques. Consequently, specialized algorithms and encoding strategies have emerged, aiming to strike a harmonious equilibrium. This paper delves into the nuanced realm of film grain handling in Versatile Video Coding (VVC) encoding. We explore the delicate balance between retaining the cinematic charm of film grain and achieving efficient compression. Moreover, we discuss the importance of perceptual quality assessment and adaptive encoding techniques in preserving film grain fidelity. Additionally, we delve into the impact of film grain handling on bitrate control and compression efficiency using VVenC, an open and optimized VVC encoder. Understanding the role of film grain and its nuanced treatment within encoders becomes increasingly pivotal for delivering high-quality, grain-inclusive content in the digital age.

Achieving reliable low-latency streaming on real-time immersive services that require seamless interaction has been of increasing importance recently. To cope with such an immersive service requirement, IETF and 3GPP defined Low Latency, Low Loss, and Scalable Throughput (L4S) architecture and terminologies to enable delay-critical applications to achieve low congestion and scalable bitrate control over 5G. With low-latency applications in mind, this paper presents a cloud-based streaming system using WebRTC for real-time communication with an adaptable L4S congestion control (aL4S-CC). aL4S-CC is designed to prevent the target service from surpassing a required end-to-end latency. It is evaluated against existing congestion controls GCC and ScreamV2 across two configurations: 1) standard L4S (sL4S) which has no knowledge of Explicit Congestion Notification (ECN) marking scheme information; 2) conscious L4S (cL4S) which recognizes the ECN marking scheme information. The results show that aL4S-CC achieves high link utilization with low latency while maintaining good performance in terms of fairness, and cL4S improves sL4S's performance by having an efficient trade-off between link utilization and latency. In the entire simulation, the gain of link utilization on cL4S is 1.4%, 4%, and 17.9% on average compared to sL4S, GCC, and ScreamV2, respectively, and the ratio of duration exceeding the target queuing delay achieves the lowest values of 1% and 0.9% for cL4S and sL4S, respectively.

This paper presents an in-depth analysis of film grain handling in open-source implementations of the Versatile Video Coding (VVC) standard. We focus on two key components: the Film Grain Analysis (FGA) module implemented in VVenC and the Film Grain Synthesis (FGS) module implemented in VVdeC. We describe the methodologies used to implement these modules and discuss the generation of Supplementary Enhancement Information (SEI) parameters to signal film grain characteristics in the encoded video sequences. Additionally, we conduct subjective and objective evaluations across Full HD videos to assess the effectiveness of film grain handling. Our results demonstrate the capability of the FGA and FGS techniques to accurately analyze and synthesize film grain, thereby improving the visual quality of encoded video content. Overall, our study contributes to advancing the understanding and implementation of film grain handling techniques in VVC open-source implementations, with implications for enhancing the viewing experience in multimedia applications.

In recent years, immersive applications have drawn the attention of industry and research. Besides, with the latest interest in the metaverse, extended reality (XR) applications are focusing on photo-realistic representations to improve the immersive experience. However, realistic representations require complex scenes and objects in terms of rendering or computation, and stand-alone or mobile XR devices still have limitations to handle such complexity. Split rendering, identified as a key enabler for 5G XR services by standardization, can address this issue by shifting complexity to the network edge, which consists of carrying out the complex rendering by powerful edge devices and transmitting a simple version of the object, such as a 2D or stereoscopic video to the end device. This article presents an implemented split rendering system designed to facilitate real-time communication while preserving the realistic representation on standalone or XR mobile devices. Moreover, to address the stringent latency requirements imposed in the split rendering, we integrate low loss, and scalable throughput (L4S) protocol with a proposed congestion control algorithm (CCA) into the split rendering system. The presented CCA demonstrates its ability to transport stable streaming while meeting the quality of service (QoS) requirements of XR services.

Video services are evolving from traditional two-dimensional video to virtual reality and holograms, which offer six degrees of freedom to users, enabling them to freely move around in a scene and change focus as desired. However, this increase in freedom translates into stringent requirements in terms of ultra-high bandwidth (in the order of Gigabits per second) and minimal latency (in the order of milliseconds). To realize such immersive services, the network transport, as well as the video representation and encoding, have to be fundamentally enhanced. The purpose of this tutorial article is to provide an elaborate introduction to the creation, streaming, and evaluation of immersive video. Moreover, it aims to provide lessons learned and to point at promising research paths to enable truly interactive immersive video applications toward holography.

We provide a mathematical framework to analyze the limits of Hybrid Automatic Repeat reQuest (HARQ) and derive analytical expressions for the most powerful test for estimating the decodability under maximum-likelihood decoding and t-error decoding. Furthermore, we numerically approximate the most powerful test for sum-product decoding. We compare the performance of previously studied HARQ prediction schemes and show that none of the state-of-the-art HARQ prediction is most powerful to estimate the decodability of a partially received signal vector under maximum-likelihood decoding and sum-product decoding. Furthermore, we demonstrate that decoding in general is suboptimal for predicting the decodability.

In recent years, applications such as cloud gaming and virtual video conferencing have gained increasing popularity and new applications, such as immersive applications, have emerged that require very low latency in order to guarantee quality of service. Such applications can benefit from using the Low Latency, Low Loss, Scalable Throughput (L4S) specification that is currently being defined by the IETF. This paper presents a congestion control algorithm aiming at achieving a target latency over a 5G connection, which together with L4S can be used in delay-critical applications. The described algorithm has been implemented into WebRTC and uses ECN marking to adapt its sending rate. The developed algorithm has been compared to the Google Congestion Control (GCC) as a baseline on various data rate patterns and shown to be more responsive to throughput variations more successfully avoiding latency spikes that surpass the acceptable latency.

In this paper, we investigate short packet transmissions in a Polar Coded Sparse Code Multiple Access (PC-SCMA) system. Since recently introduced by Arikan Polarization-Adjusted Convolutional (PAC) codes have shown remarkable performance at short block lengths with an ability of reaching theoretical limits, in this work we propose a PC-SCMA system based on the soft output PAC decoder. The proposed decoder is a combination of a sequential stack decoder and an enhanced Belief Propagation (BP) algorithm. The numerical analysis shows its ability to outperform the state-of-the-art PC-SCMA system with a Successive Cancellation List (SCL) decoder at short block lengths in terms of the bit error rate with smaller computational complexity. Furthermore, we enhance the SCL-based scheme by row-merging of the polarization kernel and evaluate its performance.

Independently coding areas in a video have potential for improved coding efficiency in Region-of-Interest (RoI) streaming applications where clients can freely switch between RoI and full view streams. This paper investigates three schemes for RoI coding in streaming scenarios based on VVC to enable harnessing open GOP coding efficiency gains. Two schemes are based on Motion-Constrained-Tile-Sets while a third scheme relies on the new VVC feature referred to as independent subpictures. The reported results show substantial bitrate gains for the RoI streams compared to naïve closed-GOP coding of up to -8.68% YUV BD-rate while for the full stream minor coding efficiency losses are reported.

In this paper, we apply an interpolation method between Reed-Muller and polar codes to investigate rate profiling for Polarization-Adjusted Convolutional (PAC) codes over the Gaussian channel at short block lengths. We evaluate a normalized compression factor to quantity the transmission efficiency of information bits for the interpolated schemes. We propose a Signal-to-Noise Ratio (SNR)-based convolutional precoding to achieve a better protection of information bits depending on the channel conditions. In the experimental part, we evaluate the new rate-profiles in compliance with the SNR-based convolutional precoding by means of sequential stack decoder and demonstrate performance-computational complexity trade-offs valuable for practical usage of PAC codes. The new optimized code with e.g. interpolated parameter equal to 0.3 brings around 50 percent of gain in terms of block error rate at high SNR values along with a slight reduction in a number of paths in stack decoder compared with a Reed-Muller baseline rate-profile.

In this work, we extend two autoencoder-based HARQ prediction schemes to exploit subcode-based features and SNR-based features jointly. We apply the proposed HARQ prediction schemes to Cloud-RAN (C-RAN) and Single-RAN (S-RAN) architectures. Furthermore, we conduct realistic link-level simulations to test the performance and compare to state-of-the-art prediction schemes that rely solely on either subcode-based features or SNR-based features. Compared to the state-of-the-art, we show that the proposed schemes reduce the transmitted redundancy at a target error rate of $5\cdot 10^{-5}$ and $2\cdot 10^{-5}$ by 12.3% - 27.3% in C-RAN architectures and 10.5% - 11.0% in S- RAN architectures, respectively.