Optical flow reliability model approximated with RBF

Authors:
Agis Rodrigo;Díaz Javier;Ortigosa Pilar;Guzmán Pablo;Ros Eduardo
Affiliations:
Computer Architecture and Electronics group, U. Almería, Almería, Spain;Department of Computer Architecture and Technology, CITIC, U. Granada, Granada, Spain;Computer Architecture and Electronics group, U. Almería, Almería, Spain;Department of Computer Architecture and Technology, CITIC, U. Granada, Granada, Spain;Department of Computer Architecture and Technology, CITIC, U. Granada, Granada, Spain
Venue:
IWANN'11 Proceedings of the 11th international conference on Artificial neural networks conference on Advances in computational intelligence - Volume Part II
Year:
2011

Citing 12
Cited 0

Universal approximation using radial-basis-function networks

Neural Computation
Performance of optical flow techniques

International Journal of Computer Vision
Rank order coding

CNS '97 Proceedings of the sixth annual conference on Computational neuroscience : trends in research, 1998: trends in research, 1998
Accuracy vs efficiency trade-offs in optical flow algorithms

Computer Vision and Image Understanding
Measurement of Image Velocity

Measurement of Image Velocity
A Statistical Confidence Measure for Optical Flows

ECCV '08 Proceedings of the 10th European Conference on Computer Vision: Part III
Continuous dimensionality characterization of image structures

Image and Vision Computing
An iterative image registration technique with an application to stereo vision

IJCAI'81 Proceedings of the 7th international joint conference on Artificial intelligence - Volume 2
Optimization strategies for high-performance computing of optical-flow in general-purpose processors

IEEE Transactions on Circuits and Systems for Video Technology
A Database and Evaluation Methodology for Optical Flow

International Journal of Computer Vision
Real-Time System for High-Image Resolution Disparity Estimation

IEEE Transactions on Image Processing
A Parallel Hardware Architecture for Scale and Rotation Invariant Feature Detection

IEEE Transactions on Circuits and Systems for Video Technology

Quantified Score

Hi-index	0.00

Visualization

Abstract

This Paper presents a new approach based on RBF NN (Radial Based Function Neural Network) in order to produce high quality optical-flow confidence estimation. The new approach is compared with a widely used confidence estimator obtaining a significant improvement. In order to evaluate the presented approach performance we have used a multi-scale version of the well known Lukas and Kanade optical flow model and widely used benchmarking optical flow sequences. The new approach aims refining optical flow representation maps but is easily applicable to other vision primitives (stereo vision, object segmentation, object recognition, object tracking, etc). Therefore, this approach represents an automatic reliability estimation model based on artificial neural networks of interest for multiple vision primitives.