3D Gaussian Splatting Rendering Performance Trade-offs on the Meta Quest 3

IEEE International Conference on Visual Communications and Image Processing (VCIP 2026)

December 13 – December 16, 2026

Singapore

[PDF]

Milad Ghanbari (AAU, Austria), Hadi Amirpour (AAU, Austria), Christian Timmerer (AAU, Austria)

Abstract: 3D Gaussian Splatting (3DGS) is a compelling technique for real-time view synthesis, but its computational demands pose significant challenges for deployment on resource-constrained standalone eXtended Reality (XR) headsets. This paper presents a systematic performance study of 3DGS rendering on the Meta Quest 3, using a custom interactive tool that enables real-time control of render resolution scale, splat density, and frame rate cap through an in-world interface. We evaluate 180 conditions formed by the factorial combination of six resolution scales (0.5–1.0), ten splat density levels (10–100%), and three camera-to-object distances (0.7 m, 1.1 m, 1.5 m), with each condition repeated ten times. Performance metrics — GPU time, frame rate, GPU utilization, CPU utilization, power draw, and stale frame count — are captured via Meta’s Oculus Virtual Reality (OVR) Metrics Tool at a fixed target frame rate of 72 Hz. The results show that resolution scale is the dominant GPU cost driver due to the quadratic growth in pixel count, while splat density reduction is only an effective performance lever when GPU time is already close to the 72 Hz frame budget. Power draw is determined by total GPU work per time unit rather than cost per frame, which creates a counterintuitive relationship where lower resolution scales draw more power by enabling higher frame rates. These findings expose the interdependence between resolution scale, splat density, frame rate, and power, and establish a quantitative foundation for rendering parameter selection in 3DGS applications on standalone XR hardware.

Bitrate Ladder Analysis for UHD Adaptive Streaming: From AVC to AV2

IEEE International Conference on Visual Communications and Image Processing (VCIP 2026)

December 13 – December 16, 2026

Singapore

[PDF]

Kamran Qureshi (AAU, Austria), Hadi Amirpour (AAU, Austria), Christian Timmerer (AAU, Austria)

Abstract: Advances in video coding continue to improve compression efficiency for adaptive bitrate (ABR) streaming. However, improved coding efficiency also changes the rate-distortion characteristics of encoded video, thereby altering the resolution transition points and the resulting bitrate ladder for UHD adaptive streaming. A convex-hull-based evaluation was performed to compare codec-dependent bitrate ladders across H.264/AVC, H.265/HEVC, H.266/VVC, AV1, and AV2 using six UHD sequences from the JVET Common Test Conditions. Experimental results showed that higher codec efficiency systematically shifts resolution transition points toward lower bitrates. Relative to AV1, AV2 reduced the median 2160p-1080p transition bitrate by approximately 3x, while the median 1080p-720p and 720p-540p transition bitrates were approximately 9x lower, allowing higher spatial resolutions to remain optimal over a wider bitrate range. Complexity analysis of the evaluated AVC, HEVC, VVC, and AV1 encoders further showed that improved compression efficiency is accompanied by increased encoding time and energy consumption. These findings demonstrate that advances in video compression influence adaptive streaming by modifying the optimal bitrate ladder rather than simply reducing bitrate requirements.

Efficient Quality Controller for Video Encoding

IEEE Visual Communications and Image Processing Conference (VCIP 2026)

December 13–16, 2026

Singapore

[PDF]

Yiying Wei (AAU, Austria), Hadi Amirpour (AAU, Austria),  and Christian Timmerer (AAU, Austria)

Abstract: Traditional video streaming relies on Adaptive Bitrate (ABR) algorithms that encode videos at fixed bitrate-resolution pairs. As a result, a rate controller is essential to ensure that each encoded representation meets its target bitrate. However, perceptually-aware bitrate ladder construction methods aim to encode videos at a fixed visual quality instead of a fixed bitrate, to avoid under- or over-allocating bits for complex and simple content. In this paper, we propose an efficient quality controller that predicts the Quantization Parameter (QP) required to achieve a target VMAF score for each video segment. The framework supports both CPU-only operation for low-complexity environments and GPU-accelerated inference for improved prediction accuracy. By leveraging content features and target quality levels, our model estimates appropriate QP values without requiring pre-encoding or tight integration with the encoder. For target VMAF scores of 94, 88, and 82, the CPU-only model achieves mean absolute errors (MAEs) of 1.05, 1.24, and 1.34, respectively, comparable to the state-of-the-art errors of 1.14, 1.27, and 1.31, while requiring only a fraction of the computational cost. The GPU-based model further reduces the MAEs to 0.50, 0.49, and 0.47, less than half of the state-of-the-art errors.

Cloud, Edge, or Split? Profiling Onboard and Split Vision-Language Model Deployment for Drone AI

Zoha Azimi, Reza Farahani, Schahram Dustdar, Christian Timmerer

The 4th International Symposium on Edge Intelligence, Trustworthy and Decentralized Artificial Intelligence (iEDGE 2026)

October 27-30, 2026 – Paris, France

Vision-Language Models (VLMs) enable edge devices like unmanned aerial vehicles (UAVs) to interpret visual observations and reason about complex environments using natural-language instructions. However, their practical deployment remains challenging as onboard inference is constrained by limited computational, memory, and energy resources, whereas cloud-based inference introduces communication latency, bandwidth overhead, and dependence on network connectivity. To address these limitations, split computing offers a promising alternative by partitioning VLM inference between the resource-constrained UAVs and more capable remote servers. However, the performance trade-offs among fully onboard, cloud-based, and split-computing architectures for lightweight VLMs have not yet been systematically profiled. This paper benchmarks these three deployment paradigms using SmolVLM-256M as a representative lightweight VLM. We quantify their inference latency, computational resource utilization, communication overhead, and energy consumption across varying image resolutions and network conditions. Our results show that no deployment strategy is universally optimal; instead, the preferred strategy depends on the interaction between network conditions and input image resolution.

Seven Years of Adaptive Streaming Research and What Comes Next

Join us as we celebrate the conclusion of the Christian Doppler Laboratory ATHENA [PDF)].

After seven years of research into adaptive video streaming and emerging networked multimedia services, we will look back at ATHENA’s achievements and explore the technologies, collaborations, and research questions shaping the future of video streaming.

Wednesday, 7 October 2026, 14:00–17:00
Alpen-Adria-Universität Klagenfurt, Stiftungssaal · Room O.0.0.1

Programme

14:00–14:05 · Welcome
14:05–14:35 · Keynote — Time to MOQ On: Leaving Legacy Latency Behind? (Ali C. Begen, Özyeğin University)
14:35–14:45 · Seven Years of ATHENA: Achievements and Impact (Christian Timmerer)
14:45–15:00 · Coffee Break
15:00–15:15 · Network-Assisted Adaptive Streaming: Toward Optimal QoE through System Collaboration (Farzad Tashtarian)
15:15–15:30 · Beyond One-Size-Fits-All: Adaptive and Computationally Efficient Video Streaming (Hadi Amirpour)
15:30–16:00 · Industry Panel: The Future of Video Streaming Chaired by Reinhard Grandl, Chief Product Officer at Bitmovin with Bitmovers and invited guests
From 16:00 · Drinks, Bites and Networking

We look forward to celebrating seven years of ATHENA with colleagues, collaborators, and friends — and to continuing the conversation about what comes next.

Free admission · Registration required via itec-sek@itec.aau.at

Interns at ATHENA (Summer 2026)

In July 2026, the ATHENA Christian Doppler Laboratory hosted four interns working on the following topics:

• Leon Kordasch – Holography
  
• Daniel Glantschnig – Automated Wind Turbine Damage Detection
• Gabriel Puri – Adaptive Streaming for Immersive Media

At the end of their internships, the interns presented their projects and findings and received official university certificates in recognition of their work. The experience proved valuable for both the interns and the ATHENA research team alike. Through personalized mentorship, hands-on training, and continuous support, the interns were able to develop strong practical skills while gaining a deeper understanding of research methodologies and technologies in the video streaming domain. We warmly thank the interns for their enthusiasm, dedication, and thoughtful feedback, which made a meaningful contribution to the ongoing work of the ATHENA lab.

Leon Kordasch: “During my internship, I worked on digital holography. I explored state-of-the-art solutions, analyzed their performance and gained lots of theoretical knowledge and technical experience. While challenging, the internship was very rewarding. My supervisor, Ayman Alkhateeb, provided guidance where needed, and collaborating with a diverse, international team made the experience both enriching and enjoyable.”

Daniel Glantschnig: “My time as an intern was both interesting and rewarding. I had the opportunity to train object detection models using datasets with and without synthetic data, then compare the results to explore how synthetic data influenced model performance. Working on this project helped me gain a much better understanding of dataset preparation, model training, evaluation, and the impact that different types of data can have on object detection systems. One of the highlights of the internship was the welcoming and friendly team, which made the experience even more enjoyable. I also greatly appreciated the support of my supervisor from the DORBINE project, Mario, who was always available to answer my questions and help me work through any challenges. The internship allowed me to apply my existing knowledge in a practical environment while developing new technical skills. Overall, I am very grateful for the opportunity, the guidance I received, and the valuable experience I gained during my time there.”

Gabriel Puri: “Over the past four weeks as an intern, I’ve had a wonderful experience. I had the opportunity to work on streaming immersive media to the Apple Vision Pro, and I even created a Swift application that streams spatial videos to a local server. This server processes the incoming stream using my integrated pipeline, which enables adaptive bitrate streaming. I’ve learned a lot about encoding and how it’s done in real-world applications such as Netflix streams. I’ve also learned through trial and error, for example by trying an approach and failing, but eventually succeeding. I also learned how to accurately grade video quality via AVQT. One of the most memorable aspects of the internship was undoubtedly the incredibly welcoming and inclusive team, as well as my supervisor and mentor, Kamran from the ATHENA project. He is a true expert in his field and provided me with crucial support. This internship has allowed me to apply my interest in computer science to useful real-world scenarios and gain industry insights. Overall, I am incredibly grateful for this opportunity and for all the guidance and support I received throughout the internship.“

Self-Training for Content-Aware Video Quality
Enhancement in HTTP Adaptive Streaming

IEEE Transactions on Broadcasting

[PDF]

Yiying Wei (AAU, Austria), Hadi Amirpour (AAU, Austria),  Wei Zhou (Cardiff University, UK), Wassim Hamidouche (TII, UAE) and Christian Timmerer (AAU, Austria)

Abstract: Fluctuations in video segment download rates and resolution switching in HTTP Adaptive Streaming (HAS) make it challenging to maintain a consistent Quality of Experience (QoE). However, the impact of such switching is often underestimated, and broadly applicable mitigation strategies remain underexplored. In the past, content-aware approaches have been introduced, using deep neural networks (DNNs) trained on an individual video segment to enhance its quality. These DNNs, transferred as a model stream alongside the video bitstream, allow clients to improve playback quality. However, transferring model streams adds bitrate overhead and additional architectural components, limiting practical use. Furthermore, supporting a wide range of device capabilities with a single DNN is impractical, as it would require device-specific models for each configuration, an approach that becomes unmanageable with increasing device heterogeneity. In this paper, we propose a new self-training method that enables clients to train content-aware video super-resolution (SR) models locally by leveraging previously downloaded high-quality segments. These segments are downscaled and used to train lightweight DNNs, which are then applied to enhance subsequent lower-quality segments. To keep training efficient and real-time, we select only a few predefined frames and extract the most informative patches using a lightweight sampling strategy. Experiments demonstrate that this approach significantly improves visual quality, with average PSNR gains of 1.07 dB (2× upscaling), 0.43 dB (3×), and 0.58 dB (4×) using ESPCN, a lightweight SR approach. To further validate the effectiveness of our approach, we conducted a series of ablation studies to analyze the contributions of individual components. Real-device measurements and end-to-end HAS simulations further show that self-training requires only 2.9–13.0% of a 4-second segment interval on CPU and 1.1–5.3% on GPU across tested mobile devices, while improving VMAF/QoE with only marginal additional rebuffering compared with generic SR.