Computation in the higher visual cortices: map-seeking circuit theory and application to machine vision

@article{Arathorn2004ComputationIT,
  title={Computation in the higher visual cortices: map-seeking circuit theory and application to machine vision},
  author={David Arathorn},
  journal={33rd Applied Imagery Pattern Recognition Workshop (AIPR'04)},
  year={2004},
  pages={73-78}
}
  • D. Arathorn
  • Published 13 October 2004
  • Computer Science
  • 33rd Applied Imagery Pattern Recognition Workshop (AIPR'04)
Map-seeking circuit theory is a biologically based computational theory of vision applicable to difficult machine vision problems such as recognition of 3D objects in arbitrary poses amid distractors and clutter, as well as to non-recognition problems such as terrain interpretation. It provides a general computational mechanism for tractable discovery of correspondences in massive transformation spaces by exploiting an ordering property of superpositions. The latter allows a set of… 
Analysis of Constrained Optimization Variants of the Map-Seeking Circuit Algorithm
TLDR
Several numerical schemes to compute local solutions are presented and compared on a pair of test problems: an image matching problem and the challenging problem of automatically solving a Rubik’s cube.
Convergence of Map Seeking Circuits
TLDR
This algorithm is formulated as discrete dynamical system on a set Δ=Πℓ=1LΔ(�), where each Δ(�¬) is a compact subset of a nonnegative orthant of ℝn, and it is shown that for an open and dense set of initial conditions in Δ the corresponding solutions converge to either a vector with unique nonzero element in each Δ (™) or to a zero vector.
Perception as an Inference Problem
TLDR
Why inferential computations are necessary for perception, and how they go beyond traditional computational approaches based on deductive processes such as feature detection and classification are described.
Detection and recognition of 3D targets in panchromatic gray scale imagery using a biologically-inspired algorithm
TLDR
This work introduces a detection stage followed by the spawning of multiple MSC processes in parallel to classify and determine the pose of the detection candidates and increases the speed of detection and allows efficient classification when multiple targets are present in the same image.
CASSIE: contextual analysis for spectral and spatial information extraction
TLDR
The research with CASSIE investigates the effectiveness of generated probability maps to reflect a human interpretation, potential benefits in terms of accuracy and processing speed for subsequent target detection, and methods for incorporating feedback from target detection algorithms to apply additional contextual constraints.
Terrain discovery and navigation of a multi-articulated linear robot using map-seeking circuits
A significant challenge in robotics is providing a robot with the ability to sense its environment and then autonomously move while accommodating obstacles. The DARPA Grand Challenge, one of the most
Questions Unanswered , and Questions Not Yet
J.M. Bower (ed.), 20 Years of Computational Neuroscience, Springer Series in Computational Neuroscience 9, DOI 10.1007/978-1-4614-1424-7_12, © Springer Science+Business Media New York 2013 Abstract I
20 Years of Learning About Vision: Questions Answered, Questions Unanswered, and Questions Not Yet Asked
I have been asked to review the progress that computational neuroscience has made over the past 20 years in understanding how vision works. In reflecting on this question, I come to the conclusion
The application of simulated annealing to a map seeking circuit
TLDR
Under these conditions, simulated annealing is shown to improve the performance of the MSC through a series of experiments involving the Hausdorff distance metric.
Instance-Based Object Recognition with Simultaneous Pose Estimation Using Keypoint Maps and Neural Dynamics
TLDR
A computationally efficient model of V1 keypoints is used to select object parts with the highest information content and model their surroundings using simple colour features and is fed into a dynamical neural network which performs pose, scale and translation estimation of the object given a set of previously observed object views.
...
1
2
...

References

SHOWING 1-9 OF 9 REFERENCES
Map-Seeking Circuits in Visual Cognition: A Computational Mechanism for Biological and Machine Vision
TLDR
The author argues that map-seeking is a fundamental, broadly applicable computational operation with algorithmic, neuronal, and analog electronic implementations, and that its generality makes it suitable as the core of a computational explanation for several cognitive functions.
Emergence of simple-cell receptive field properties by learning a sparse code for natural images
TLDR
It is shown that a learning algorithm that attempts to find sparse linear codes for natural scenes will develop a complete family of localized, oriented, bandpass receptive fields, similar to those found in the primary visual cortex.
Circuits for Local and Global Signal Integration in Primary Visual Cortex
TLDR
It is found that monosynaptic horizontal connections within area V1 are of an appropriate spatial scale to mediate interactions within the SF of V1 neurons and to underlie contrast-dependent changes in SF size, which could represent an anatomical substrate for contextual modulation and global-to-local integration of visual signals.
Recognition under transformation using superposition ordering property
A mathematical property of pattern superpositions permits the construction of mechanisms capable of recognising learned patterns occurring under visual transformation within a visually complex field.
Computational subunits in thin dendrites of pyramidal cells
TLDR
This work combined confocal imaging and dual-site focal synaptic stimulation of identified thin dendrites in rat neocortical pyramidal neurons found that nearby inputs on the same branch summed sigmoidally, whereas widely separated inputs or inputs to different branches summed linearly.
A Solution to the Generalized Correspondence Problem Using an Ordering Property of Superpositions
  • 2004
Map - Seeking : Recognition Under Transformation Using A Superposition Ordering Property
  • Electronics Letters
  • 2001
Map-Seeking: Recognition Under Transformation Using A Superposition Ordering Property
  • Electronics Letters
  • 2001
A Solution to the Generalized Correspondence Problem Using an Ordering Property of Superpositions, submitted 2004. 3D model of T-80 and WW-2 tank courtesy 3DCafe
  • com Proceedings of the 33rd Applied Imagery Pattern Recognition Workshop (AIPR'04)