Object Class Recognition and Localization Using Sparse Features with Limited Receptive Fields

Mutch, Jim; Lowe, David

doi:10.1007/s11263-007-0118-0

Cited by 321 publications

(342 citation statements)

References 32 publications

(60 reference statements)

Supporting

Mentioning

337

Contrasting

Unclassified

Order By: Relevance

“…We assume that the hierarchical processing in both pathways is organized in a similar fashion and, thus, make use of a generic processing architecture for neural feature extraction as shown in Figure 1. The architecture is a modified variant of the object-recognition model proposed by [10,16,26]. In layer S1, different scale representations of the input image are convolved with 2D Gabor filters of different orientations (form path) and a spatiotemporal correlation detector is used to build a discrete velocity space representation (motion path).…”

Section: Visual Featuresmentioning

confidence: 99%

“…The layer S2 is created by a simple template matching of patches of C1 activities against a number of prototype patches. These prototypes are randomly selected during the learning stage (for details, see [10]). In the final layer C2, the S2 prototype responses are again pooled over a limited neighborhood and combined into a single feature vector which serves as input to the successive classification stage.…”

Section: Visual Featuresmentioning

confidence: 99%

“…The variables x and y range from −5 to 5 and θ varies between 0 and π in steps of π/6. The aspect ratio γ, the effective width σ, and the wavelength λ are set to 0.3, 4.5, and 5.6, respectively (in accordance with [10]. This results in six orientation maps at each scale.…”

Section: Visual Featuresmentioning

confidence: 99%

See 2 more Smart Citations

Multiple Classifier Systems for the Classification of Audio-Visual Emotional States

Glodek

Tschechne

Layher

et al. 2011

Lecture Notes in Computer Science

103

View full text Add to dashboard Cite

Abstract. Research activities in the field of human-computer interaction increasingly addressed the aspect of integrating some type of emotional intelligence. Human emotions are expressed through different modalities such as speech, facial expressions, hand or body gestures, and therefore the classification of human emotions should be considered as a multimodal pattern recognition problem. The aim of our paper is to investigate multiple classifier systems utilizing audio and visual features to classify human emotional states. For that a variety of features have been derived. From the audio signal the fundamental frequency, LPCand MFCC coefficients, and RASTA-PLP have been used. In addition to that two types of visual features have been computed, namely form and motion features of intermediate complexity. The numerical evaluation has been performed on the four emotional labels Arousal, Expectancy, Power, Valence as defined in the AVEC data set. As classifier architectures multiple classifier systems are applied, these have been proven to be accurate and robust against missing and noisy data.

show abstract

Section: Visual Featuresmentioning

confidence: 99%

Section: Visual Featuresmentioning

confidence: 99%

See 1 more Smart Citation

Multiple Classifier Systems for the Classification of Audio-Visual Emotional States

Glodek

Tschechne

Layher

et al. 2011

Lecture Notes in Computer Science

103

View full text Add to dashboard Cite

show abstract

“…These findings in conjunction with experimental studies of the visual cortex justify the use of such filters in the so-called standard model for object recognition (Riesenhuber & Poggio, 1999;Serre et al, 2005;Mutch & Lowe, 2008), whose filters are fixed, in contrast to those of Convolutional Neural Networks (CNNs) (LeCun et al, 1998;Behnke, 2003;Simard et al, 2003), whose weights (filters) are randomly initialized and learned in a supervised way using back-propagation (BP). A DNN, the basic building block of our proposed MCDNN, is a hierarchical deep neural network, alternating convolutional with max-pooling layers (Riesenhuber & Poggio, 1999;Serre et al, 2005;Scherer et al, 2010).…”

Section: Introductionmentioning

confidence: 97%

Multi-column deep neural network for traffic sign classification

et al. 2012

View full text Add to dashboard Cite

We describe the approach that won the final phase of the German traffic sign recognition benchmark. Our method is the only one that achieved a betterthan-human recognition rate of 99.46%. We use a fast, fully parameterizable GPU implementation of a Deep Neural Network (DNN) that does not require careful design of pre-wired feature extractors, which are rather learned in a supervised way. Combining various DNNs trained on differently preprocessed data into a Multi-Column DNN (MCDNN) further boosts recognition performance, making the system insensitive also to variations in contrast and illumination.

show abstract

“…The local contour information that is extracted by orientation-selective cells provides basis for further, more complex visual tasks, such as object recognition [16,14,12,19,17,15]. The performance of various contour operators, which are inspired by the function of simple cells, in contour detection tasks has, however, not been quantified and they have not been compared in that respect.…”

Section: Introductionmentioning

confidence: 99%

Contour Detection by CORF Operator

Azzopardi

Petkov

2012

Artificial Neural Networks and Machine Learning – ICANN 2012

View full text Add to dashboard Cite

Abstract. We propose a contour operator, called CORF, inspired by the properties of simple cells in visual cortex. It combines, by a weighted geometric mean, the blurred responses of difference-of-Gaussian operators that model cells in the lateral geniculate nucleus (LGN). An operator that has gained particular popularity as a computational model of a simple cell is based on a family of Gabor Functions (GFs). However, the GF operator short-cuts the LGN, and its effectiveness in contour detection tasks, which is assumed to be the primary biological role of simple cells, has never been compared with the effectiveness of alternative operators. We compare the performances of the CORF and the GF operators using the RuG and the Berkeley data sets of natural scenes with associated ground truths. The proposed CORF operator outperforms the GF operator (RuG: t(39) = 4.39, p < 10 −4 and Berkeley: t(499) = 4.95, p < 10 −6 ).

show abstract

Object Class Recognition and Localization Using Sparse Features with Limited Receptive Fields

Abstract: We

Cited by 321 publications

References 32 publications

Multiple Classifier Systems for the Classification of Audio-Visual Emotional States

Multiple Classifier Systems for the Classification of Audio-Visual Emotional States

Multi-column deep neural network for traffic sign classification

Contour Detection by CORF Operator

Contact Info

Product

Resources

About