Belief Propagation Implementation Using CUDA on an NVIDIA GTX 280

@inproceedings{Xu2009BeliefPI,
  title={Belief Propagation Implementation Using CUDA on an NVIDIA GTX 280},
  author={Yanyan Xu and Hui Chen and Reinhard Klette and Jiaju Liu and Tobi Vaudrey},
  booktitle={Australasian Conference on Artificial Intelligence},
  year={2009}
}
Disparity map generation is a significant component of vision-based driver assistance systems. This paper describes an efficient implementation of a belief propagation algorithm on a graphics card (GPU) using CUDA (Compute Uniform Device Architecture) that can be used to speed up stereo image processing by between 30 and 250 times. For evaluation purposes, different kinds of images have been used: reference images from the Middlebury stereo website, and real-world stereo sequences, self… 

CUDA implementation of belief propagation for stereo vision

  • Young-kyu Choi
  • Computer Science
    13th International IEEE Conference on Intelligent Transportation Systems
  • 2010
TLDR
This paper provides profiling result to find the performance bottleneck of BP, and proposes a new parallelization technique to speed up the message computation, which takes up the longest time in BP.

Architectures for Stereo Vision

TLDR
The theoretic background on stereo vision systems is provided, architectures and implementations for real-time applications for stereo matching are discussed, and the major differences in designing parallelization techniques for extremely different image processing platforms are illustrated.

Efficient parallel message computation for MAP inference

TLDR
This paper revisits message computation as a distance transformation using a more formal setting than [8] to generalize it to arbitrary PPFs and provides results as accurate as state of the art inference methods, though is in general one order of magnitude faster in terms of speed.

Design of Interpolation Functions for Subpixel-Accuracy Stereo-Vision Systems

TLDR
This work proposes two methodologies for interpolation function generation based on data generated by the stereo algorithms that employs synthetic images of a known environment and applies function fitting to the resulted data.

GPUMLib : An Efficient Open-Source GPU Machine Learning Library

TLDR
This work proposes the creation of an open source GPU Machine Learning Library (GPUMLib) that aims to provide the building blocks for the development of efficient GPU ML software and shows that the algorithms already implemented yield significant time savings over the CPU counterparts.

Accelerating Genetic Programming through Graphics Processing Units.

TLDR
In this chapter, Genetic Programming methods currently ported are surveyed and the true degree of parallelism of GPUs is often hidden from the user, making programming even more flexible and convenient.

Detecting, segmenting and tracking unknown objects using multi-label MRF inference

Fusion of 3D-LIDAR and camera data for scene parsing

GPU-Accelerated Clique Tree Propagation for Pouch Latent Tree Models

TLDR
This paper considers using GPUs to exploit two parallelism opportunities, namely data parallelism and element-wise parallelism, for PTLMs and focuses on clique tree propagation, since this exact inference procedure is a strenuous task and is recurrently called for each data sample and each model structure during PLTM training.

A GPGPU-Based Acceleration of Fault-Tolerant MLP Learnings

TLDR
A method to speed-up the learning processes proposed in [1] using the GPGPU technology is proposed, and a multilayer perceptron (MLP) which has successfully trained using the Deep LM is tolerant to multiple weight and neuron faults.

References

SHOWING 1-10 OF 16 REFERENCES

Belief Propagation on the GPU for Stereo Vision

TLDR
A novel implementation of Bayesian belief propagation for graphics processing units found in most modern desktop and notebook computers is presented, and applies it to the stereo problem.

GPU implementation of belief propagation using CUDA for cloud tracking and reconstruction

TLDR
An efficient CUDA-based GPU implementation of the belief propagation algorithm that can be used to speed up stereo image processing and motion tracking calculations without loss of accuracy is described.

Using graphics devices in reverse: GPU-based Image Processing and Computer Vision

  • J. FungSteve Mann
  • Computer Science
    2008 IEEE International Conference on Multimedia and Expo
  • 2008
TLDR
This paper discusses how this processing power is being harnessed for image processing and computer vision, thereby providing dramatic speedups on commodity, readily available graphics hardware.

OpenVIDIA: parallel GPU computer vision

TLDR
This paper proposes using GPUs in approximately the reverse way: to assist in "converting pictures into numbers" (i.e. computer vision) and provides a simple API which implements some common computer vision algorithms.

Real-time Global Stereo Matching Using Hierarchical Belief Propagation

TLDR
This paper presents a belief propagation based global algorithm that generates high quality results while maintaining real-time performance, and is the first BP based global method that runs at real- time speed.

Feature tracking and matching in video using programmable graphics hardware

TLDR
Novel implementations of the KLT feature tracking and SIFT feature extraction algorithms that run on the graphics processing unit (GPU) and is suitable for video analysis in real-time vision systems are described.

Efficient Belief Propagation for Early Vision

TLDR
Algorithmic techniques are presented that substantially improve the running time of the loopy belief propagation approach and reduce the complexity of the inference algorithm to be linear rather than quadratic in the number of possible labels for each pixel, which is important for problems such as image restoration that have a large label set.

CUDA cuts: Fast graph cuts on the GPU

  • Vibhav VineetP. Narayanan
  • Computer Science
    2008 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
  • 2008
TLDR
This paper presents an implementation of the push-relabel algorithm for graph cuts on the GPU that can perform over 60 graph cuts per second on 1024times1024 images and over 150 graph cutsper second on 640times480 images on an Nvidia 8800 GTX.

A Taxonomy and Evaluation of Dense Two-Frame Stereo Correspondence Algorithms

TLDR
This paper has designed a stand-alone, flexible C++ implementation that enables the evaluation of individual components and that can easily be extended to include new algorithms.

Belief Propagation for Stereo Analysis of Night-Vision Sequences

TLDR
This paper considers other options of preprocessing (Canny and Kovesi-Owens edge operators), and concludes with a recommended setting for belief propagation on those sequences.