Fast stereoscopic rendering on mobile ray tracing GPU for virtual reality applications

Lee, Won-Jong; Hwang, Seok Joong; Shin, Youngsam; Yoo, Jeong-Joon; Ryu, Soojung

doi:10.1109/icce.2017.7889353

Cited by 7 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The scene's AABB is subdivided into cubic cells whose size is chosen to be as fine as possible, while the number of geometry cells, which intersect with an object ("# of geom. cells"), cannot exceed 65,536 (=2 16 ). In the table, "Cell resolution" shows how finely the scene space is tessellated in cubic cells.…”

Section: Quantization Of Vertex Attributesmentioning

confidence: 98%

“…Modern high-end PC GPUs often enable real-time ray tracing for nontrivial 3D scenes [5]. On the other hand, despite the endeavors dedicated to developing a mobile accelerator for ray tracing [6,[15][16][17], a single modern mobile GPU still does not provide enough performance for real-time rendering of large 3D scenes.…”

Section: Related Workmentioning

confidence: 99%

“…The first step toward our space-efficient mesh representation is to partition the (slightly enlarged) axis-aligned bounding box (AABB) of a given scene into equal-sized cubic cells, and find those cells, called the geometry cells, that are intersected by the scene's triangular mesh (see Figure 1a). A key factor in partitioning the scene space is to set the size of the cubic cell to be as small as possible while the resulting number of geometry cells cannot exceed 65,536 (=2 16 ). This can be done by subdividing the longest side of the AABB repeatedly, in which the desired number of subintervals is decided in a bisection manner from an initial range, say, [2 5 , 2 10 ].…”

Section: Quantization Of Vertex Attributesmentioning

confidence: 99%

“…Contrary to the vertex position each component of which is quantized within a cell, the vertex normal is quantized as a whole. For run-time access during rendering, we adopt a simple table-based quantization technique where a preprocessed lookup table, indexed with a 16-bit unsigned integer, provides up to 65,536 (=2 16 ) discrete unit normal samples. To fill the table, we use the normal encoding/decoding method of Smith et al [24], which exhibits the lowest mean and maximum angular errors among a variety of 16-bit unit normal representation techniques [25].…”

Section: Quantization Of Vertex Attributesmentioning

confidence: 99%

“…Note that for a triangular mesh in the indexed-face-set representation with n v vertices and n f faces, the original mesh dataset actually needs 32n v + 12n f + 8 bytes of memory, among which the 8 additional bytes are for storing the values of n v and n f (a four-byte unsigned int each). On the other hand, the presented quantization method requires 12n v + 12n f bytes (for the lists of quantized vertex attributes and triangle vertex indices) plus 3 • 2 • 2 16 bytes (for the cell index table in which each index component is saved as a two-byte short integer), 3 • 4 • 2 16 bytes (for the normal lookup table in which each normal component is saved as a four-byte float) and 24 bytes (for the values of n v (4), n f (4), the xyz coordinates of AABB's anchor point (12), and the cell size( 4)). This leads to the size reduction ratio of…”

Section: Analysis Of Triangular-mesh Compressionmentioning

confidence: 99%

See 4 more Smart Citations

Efficient Ray Tracing of Large 3D Scenes for Mobile Distributed Computing Environments

Seo

Park

Ihm

2022

Sensors

View full text Add to dashboard Cite

Cluster computing has attracted much attention as an effective way of solving large-scale problems. However, only a few attempts have been made to explore mobile computing clusters that can be easily built using commodity smartphones and tablets. To investigate the possibility of mobile cluster-based rendering of large datasets, we developed a mobile GPU ray tracer that renders nontrivial 3D scenes with many millions of triangles at an interactive frame rate on a small-scale mobile cluster. To cope with the limited processing power and memory space, we first present an effective 3D scene representation scheme suitable for mobile GPU rendering. Then, to avoid performance impairment caused by the high latency and low bandwidth of mobile networks, we propose using a static load balancing strategy, which we found to be more appropriate for the vulnerable mobile clustering environment than a dynamic strategy. Our mobile distributed rendering system achieved a few frames per second when ray tracing 1024 × 1024 images, using only 16 low-end smartphones, for large 3D scenes, some with more than 10 million triangles. Through a conceptual demonstration, we also show that the presented rendering scheme can be effectively explored for augmenting real scene images, captured or perceived by augmented and mixed reality devices, with high quality ray-traced images.

show abstract

Section: Quantization Of Vertex Attributesmentioning

confidence: 98%

Section: Related Workmentioning

confidence: 99%

Section: Quantization Of Vertex Attributesmentioning

confidence: 99%

Section: Quantization Of Vertex Attributesmentioning

confidence: 99%

Section: Analysis Of Triangular-mesh Compressionmentioning

confidence: 99%

See 3 more Smart Citations

Efficient Ray Tracing of Large 3D Scenes for Mobile Distributed Computing Environments

Seo

Park

Ihm

2022

Sensors

View full text Add to dashboard Cite

show abstract

What We Know About the Greenability of Reality Technologies: A Systematic Literature Review

Khakpour

Sánchez‐Gordón

Colomo‐Palacios

2020

Contributions to Management Science

View full text Add to dashboard Cite

Strategies for the Digitalization of Cultural Heritage

Gervasi

Perri

Simonetti

et al. 2022

Lecture Notes in Computer Science

View full text Add to dashboard Cite

The COVID-19 pandemic caused despair, poverty and, above all, pain and death across the planet. Nevertheless, there is no doubt that it has also given a strong global impetus to the digital world, asserting its importance, sustainability and richness of perspectives. With our work we intend to establish a set of best practices aimed at defining methods and technologies that will enable those fascinated by digital technologies to contribute effectively to the digitisation of cultural heritage on a large scale. Various aspects that play a crucial role in the digitisation of artifacts will be discussed, with a focus on the issues involved in the manual realisation of works using Blender and Unity software. For demonstration purposes, two very popular use cases in the Umbria region of Italy are presented: “Piazza IV Novembre” in Perugia with the magnificent “Fontana Maggiore”, the “Palazzo dei Priori” and the Duomo on one side and the Republic square in Foligno on the other, with the Duomo, the Bishop’s house and the Diocesan Museum, the Town Hall and “Palazzo Trinci”. The first realization was carried out using photogrammetry techniques, the software Blender and Unity, while the second was carried out exclusively with Blender and Unity.The theme is highly relevant in Europe, particularly Italy, where the topic is part of the post-COVID-19 National Recovery and Resilience Plan (NRRP).

show abstract

Fast stereoscopic rendering on mobile ray tracing GPU for virtual reality applications

Cited by 7 publications

References 6 publications

Efficient Ray Tracing of Large 3D Scenes for Mobile Distributed Computing Environments

Efficient Ray Tracing of Large 3D Scenes for Mobile Distributed Computing Environments

What We Know About the Greenability of Reality Technologies: A Systematic Literature Review

Strategies for the Digitalization of Cultural Heritage

Contact Info

Product

Resources

About