The methods for visual learning that compure a space of eigenvectors by Principal Component Ariulysis (PCA) rradirionall? require a batch computation step. Since rhis leads to potential problems when dealing wirh large sets of image.r, several incremental merhods for the cornputation of the eigenvectors huve been introduced. Howeve,: such learning curinof be considered as an oil-line pmcess, since all rlie images are rerained uriril the final step of computation of space of eigetivecrors, when their coeffcienrs in rhis sirbspuce ure comprired. I n this paper we propose a method that allows for simulraneous leurning and recognition. We show rhur we cuii keep only rhe coefficients of the learned images and discard rhe acriral images and still ure able ro build a model of appearance rhar is fast ro compicre mid open-ended. We performed exrensive erperimenral tesring which showed that the recognirion rare and reconstruction accuracy are comparable to those obrained by the batch method.

Abstract-When using appearance-based recognition for self-localization of mobile robots, the images obtained during the exploration of the environment need to be efficiently stored in the memory. PCA offers means for representing the images in a low-dimensional subspace, which allows for efficient matching and recognition. For active exploration it is necessary to use an incremental method for the computation of the subspace. While such methods have been considered before, only the on-line construction of eigenvectors has been addressed. Representations of the images in the subspace were computed only after the final subspace had been built, requiring that all the images were kept in the memory. In this paper we propose to use an incremental PCA algorithm with the updating of partial image representations in a way that allows the robot to discard the acquired images immediately after the update. Such a model is open-ended, meaning that we can easily update it with new images. We show that the performance of the proposed method is comparable to the performance of the batch method in terms of compression, computational cost and the precision of localization. We also show that by applying the repetitive learning, the subspace converges to that constructed with the batch method.Keywords-Robot localization, on-line visual learning, PCA updating, view-based robot localization, repetitive learning.

Perception is often biased by secondary stimulus attributes (e.g., stimulus noise, attention, or spatial context). A correct quantitative characterization of perceptual bias is essential for testing hypotheses about the underlying perceptual mechanisms and computations. We demonstrate that the standard two-alternative forced choice (2AFC) method can lead to incorrect estimates of perceptual bias. We present a new 2AFC method that solves this problem by asking subjects to judge the relative perceptual distances between the test and each of two reference stimuli. Naïve subjects can easily perform this task. We successfully validated the new method with a visual motion-discrimination experiment. We demonstrate that the method permits an efficient and accurate characterization of perceptual bias and simultaneously provides measures of discriminability for both the reference and test stimulus, all from a single stimulus condition. This makes it an attractive choice for the characterization of perceptual bias and discriminability in a wide variety of psychophysical experiments.

Object motion in natural scenes results in visual stimuli with a rich and broad spatiotemporal frequency spectrum. While the question of how the visual system detects and senses motion energies at different spatial and temporal frequencies has been fairly well studied, it is unclear how the visual system integrates this information to form coherent percepts of object motion. We applied a combination of tailored psychophysical experiments and predictive modeling to address this question with regard to perceived motion in a given direction (i.e., stimulus speed). We tested human subjects in a discrimination experiment using stimuli that selectively targeted four distinct spatiotemporally tuned channels with center frequencies consistent with a common speed. We first characterized subjects' responses to stimuli that targeted only individual channels. Based on these measurements, we then predicted subjects' psychometric functions for stimuli that targeted multiple channels simultaneously. Specifically, we compared predictions of three Bayesian observer models that either optimally integrated the information across all spatiotemporal channels, or only used information from the most reliable channel, or formed an average percept across channels. Only the model with optimal integration was successful in accounting for the data. Furthermore, the proposed channel model provides an intuitive explanation for the previously reported spatial frequency dependence of perceived speed of coherent object motion. Finally, our findings indicate that a prior expectation for slow speeds is added to the inference process only after the sensory information is combined and integrated.

Abstract

No abstract

No abstract

In this paper, we propose a novel method for efficiently calculating the eigenvectors of uniformly rotated images of a set of templates. As we show, the images can be optimally approximated by a linear series of eigenvectors which can be calculated without actually decomposing the sample covariance matrix.