Amortized supersampling

Yang, Lei; Nehab, Diego; Sander, Pedro V.; Sitthi‐Amorn, Pitchaya; Lawrence, Jason; Hoppe, Hugues

doi:10.1145/1618452.1618481

Cited by 61 publications

(68 citation statements)

References 36 publications

(17 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In most cases we limit our discussion to results that are directly relevant to application decisions. For the detailed derivation of these results, please refer to Sitthi-amorn et al (2008a,b); Yang et al (2009).…”

Section: Discussionmentioning

confidence: 99%

“…Although obtaining a closed-form expression for σ 2 t (p) is impractical for arbitrary scene motion (with v changing every frame), the case of constant panning motion with fixed v is tractable and can be used to locally approximate other types of motion. Following the derivations from Yang et al (2009) without applying the exponential smoothing, the variance of blur σ after n frames of reprojection is:…”

Section: Resampling Errormentioning

confidence: 99%

“…Yang et al (2009) propose a scheme to extend the technique to approximate subpixel accuracy. Here we give a brief description of the underlying ideas.…”

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

“…Conceptually, we forward reproject all the previous samples into the current Figure 15: Sampling from multiple subpixel buffers. To properly reconstruct the quadrant value, Yang et al (2009) use nonuniform blending weights defined by a tent function centered on the quadrant being updated. (a) In the absence of local motion, only the correct pixel has non-zero weight in the tent, so no resampling blur is introduced; (b) For a moving scene, the samples are weighted using the tent function, and higher weights are given to samples closer to the desired quadrant center to limit the amount of blur.…”

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

“…Yang et al (2009) propose methods to estimate the reconstruction error, as well as the signal change in real-time, and limit α accordingly such that a minimum amount of refresh is guaranteed. The reconstruction error is estimated by deriving an empirical relationship between the fractional pixel velocity v, α and the error, similar to the purpose of Equation 4.…”

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

See 4 more Smart Citations

Exploiting temporal coherence in real-time rendering

Scherzer

Yang

Mattausch

2010

ACM SIGGRAPH ASIA 2010 Courses

Self Cite

View full text Add to dashboard Cite

Temporal coherence (TC), the correlation of contents between adjacent rendered frames, exists across a wide range of scenes and motion types in practical real-time rendering. By taking advantage of TC, we can save redundant computation and improve the performance of many rendering tasks significantly with only a marginal decrease in quality. This not only allows us to incorporate more computationally intensive shading effects to existing applications, but also offers exciting opportunities of extending high-end graphics applications to reach lower-spec consumer-level hardware.This course aims to introduce participants to the concepts of TC, and provide them the working practical and theoretical knowledge to exploit TC in a variety of shading tasks. It begins with an introduction of the general notion of TC in rendering, as well as an overview of the recent developments in this field. Then it focuses on a key data structure -the reverse reprojection cache, which is the foundation of many applications. The course proceeds with a number of extensions of the basic algorithm for assisting in multi-pass shading effects, shader antialiasing, casting shadows and globalillumination effects. Finally, several more general coherence topics beyond pixel reuse are introduced, including visibility culling optimization and object-space global-illumination approximations. For all the major techniques and applications covered, implementation and practical issues involved in development are addressed in detail.In general, we emphasize "know how" and the guidelines related to algorithm choices. After the course, participants are encouraged to find and utilize TC in their own applications and rapidly adapt existing algorithms to meet their requirements.The version of the course notes you are currently reading was created at September 23, 2010. The newest version of these course notes can be downloaded from this URL. Course PrerequisitesThis course should give a computer graphics practitioner, student or researcher the working and theoretical knowledge to implement state-of-the-art temporal coherence techniques. The content of this course is delivered by talks and course notes. Prerequisites incldue knowledge of basic real-time computer graphics, such as programmable shading pipeline, model transformation, rasterization and texture mapping. Experience in writing vertex and fragment shaders is preferred.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Resampling Errormentioning

confidence: 99%

“…Yang et al (2009) propose a scheme to extend the technique to approximate subpixel accuracy. Here we give a brief description of the underlying ideas.…”

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

Section: Achieving Subpixel Accuracy: Shader Antialiasingmentioning

confidence: 99%

See 3 more Smart Citations

Exploiting temporal coherence in real-time rendering

Scherzer

Yang

Mattausch

2010

ACM SIGGRAPH ASIA 2010 Courses

Self Cite

View full text Add to dashboard Cite

show abstract

Neural foveated super‐resolution for real‐time VR rendering

Ye,

Meng,

Guo

et al. 2024

Computer Animation & Virtual

View full text Add to dashboard Cite

As virtual reality display technologies advance, resolutions and refresh rates continue to approach human perceptual limits, presenting a challenge for real‐time rendering algorithms. Neural super‐resolution is promising in reducing the computation cost and boosting the visual experience by scaling up low‐resolution renderings. However, the added workload of running neural networks cannot be neglected. In this article, we try to alleviate the burden by exploiting the foveated nature of the human visual system, in a way that we upscale the coarse input in a heterogeneous manner instead of uniform super‐resolution according to the visual acuity decreasing rapidly from the focal point to the periphery. With the help of dynamic and geometric information (i.e., pixel‐wise motion vectors, depth, and camera transformation) available inherently in the real‐time rendering content, we propose a neural accumulator to effectively aggregate the amortizedly rendered low‐resolution visual information from frame to frame recurrently. By leveraging a partition‐assemble scheme, we use a neural super‐resolution module to upsample the low‐resolution image tiles to different qualities according to their perceptual importance and reconstruct the final output adaptively. Perceptually high‐fidelity foveated high‐resolution frames are generated in real‐time, surpassing the quality of other foveated super‐resolution methods.

show abstract

Improving Temporal Antialiasing with Adaptive Ray Tracing

et al. 2019

View full text Add to dashboard Cite

In this chapter, we discuss a pragmatic approach to real-time supersampling that extends commonly used temporal antialiasing techniques with adaptive ray tracing. The algorithm conforms to the constraints of a commercial game engine, removes blurring and ghosting artifacts associated with standard temporal antialiasing, and achieves quality approaching 16× supersampling of geometry, shading, and materials within the 16 ms frame budget required of most games.

show abstract

Amortized supersampling

Cited by 61 publications

References 36 publications

Exploiting temporal coherence in real-time rendering

Exploiting temporal coherence in real-time rendering

Neural foveated super‐resolution for real‐time VR rendering

Improving Temporal Antialiasing with Adaptive Ray Tracing

Contact Info

Product

Resources

About