We present an algorithm for the on-board vision vehicle detection problem using a cascade of boosted classifiers. Three families of features are compared: the rectangular filters (Haar-like features), the histograms of oriented gradient (HoG), and their combination (a concatenation of the two preceding features). A comparative study of the results of the generative (HoG features), discriminative (Haar-like features) detectors, and of their fusion is presented. These results show that the fusion combines the advantages of the other two detectors: generative classifiers eliminate "easily" negative examples in the early layers of the cascade, while in the later layers, the discriminative classifiers generate a fine decision boundary removing the negative examples near the vehicle model. The best algorithm achieves good performances on a test set containing some 500 vehicle images: the detection rate is about 94% and the false-alarm rate per image is 0.0003.

https://link.springer.com/content/pdf/10.1155%2F2008%2F782432.pdf

A cascade of boosted generative and discriminative classifiers for vehicle detection

https://hal.archives-ouvertes.fr/hal-00668271/document

Off-line handwritten word recognition using multi-stream hidden Markov models

On-line handwritten digit recognition based on trajectory and velocity modeling

This paper describes an application of the minimum classification error (MCE) criterion to the problem of recognizing online unconstrained-style characters and words. We describe an HMM-based, character and word-level MCE training aimed at minimizing the character or word error rate while enabling flexibility in writing style through the use of multiple allographs per character. Experiments on a writer-independent character recognition task covering alpha-numerical characters and keyboard symbols show that the MCE criterion achieves more than 30 percent character error rate reduction compared to the baseline maximum likelihood-based system. Word recognition results, on vocabularies of 5k to 10k, show that MCE training achieves around 17 percent word error rate reduction when compared to the baseline maximum likelihood system

Minimum classification error training for online handwriting recognition

This paper introduces an event-based luminance-free feature from the output of asynchronous event-based neuromorphic retinas. The feature consists in mapping into a matrix, the distribution of the optical flow along the contours of the moving objects in the visual scene. Asynchronous event-based neuromorphic retinas are composed of autonomous pixels, each of them asynchronously generating ''spiking'' events that encode relative changes in pixels' illumination at high temporal resolutions. The optical flow is computed at each event, and is integrated locally or globally in a speed and direction coordinate frame based grid, using speed-tuned temporal kernels. The latter ensures that the resulting feature represents equitably the distribution of the normal motion along the current moving edges, whatever their respective dynamics. The usefulness and the genericness of the proposed feature are demonstrated in pattern recognition applications: local corner detection and global gesture recognition.

/pdf/a-motion-based-feature-for-event-based-pattern-recognition-1oxnawkqi5.pdf

A Motion-Based Feature for Event-Based Pattern Recognition.

Handwriting recognition is such a complex classification problem that it is quite usual now to make co-operate several classification methods at the pre-processing stage or at the classification stage. In this paper, we present an original two stages recognizer. The first stage is a model-based classifier that stores an exhaustive set of character models. The second stage is a discriminative classifier that separates the most ambiguous pairs of classes. This hybrid architecture is based on the idea that the correct class almost systematically belongs to the two more relevant classes found by the first classifier. Experiments on the Unipen database show a 30% improvement on a 62 class recognition problem.

Combining model-based and discriminative classifiers: application to handwritten character recognition

Handwriting recognition for hand-held devices like PDAs requires very accurate and adaptive classifiers. It is such a complex classification problem that it is quite usual now to make co-operate several classification methods. In this paper, we present an original two stages recognizer. The first stage is a model-based classifier which store an exhaustive set of character models. The second stage is a pairwise classifier which separate the most ambiguous pairs of classes. This hybrid architecture is based on the idea that the correct class almost systematically belongs to the two more relevant classes found by the first classifier. Experiments on a 80,000 examples database show a 30% improvement on a 62 classes recognition problem. Moreover, we show experimentally that such an architecture suits perfectly for incremental classification.

Hybrid generative/discriminative classifier for unconstrained character recognition

We recently designed a hand-printed text recognizer. The system is based on three set of experts respectively used to segment, classify and validate the text (with a French lexicon : 200K words). We present in this communication writer adaptation methods. The first is supervised by the user. The others are self-supervised strategies which compare classification hypothesis with lexical hypothesis and modify consequently classifier parameters. The last method increases the system accuracy and the classification speed. Experiments are presented on a large database of 90 texts (5400 words) written by 54 different writers and good recognition rates (82%) have been obtained.

Self-supervised writer adaptation using perceptive concepts : application to on-line text recognition

We developed a handwritten text recognizer for on-line text written on a touch-terminal. This system is based on the activation-verification cognitive model. It is composed of three experts dedicated respectively to signal segmentation in symbols, symbol classification and lexical analysis of the classification results. The baseline system is writer-independent. We present in this paper several strategies of self-supervised writer-adaptation that we compare to the supervised adaptation scheme. The best strategy called "prototype dynamic management" modifies the recognizer parameters allowing to get results close to the supervised methods. Results are presented on a 90 texts (5400 words) database written by 38 different writers.

/pdf/self-supervised-adaptation-for-on-line-text-recognition-3m96vcxwcd.pdf

Self-supervised adaptation for on-line text recognition

Nous avons recemment developpe un systeme complet de reconnaissance de textes manuscrits saisis sur tablette a digitaliser. Ce systeme se base sur le modele cognitif d'activation verification et s'articule autour de trois experts dedies respectivement a la segmentation du signal ecrit en symboles, la classification des symboles et l'analyse lexicale des resultats de classification. Nous presentons dans cet article plusieurs strategies d'adaptation non-supervisees au scripteur que nous comparons aux methodes d'adaptation supervisees. Nous proposons une strategie de modification dynamique non-supervisee des parametres du classifieur permettant d'obtenir des resultats proches des methodes supervisees. Les resultats obtenus sur une base de donnees de 90 textes (5 400 mots) ecrits par 38 scripteurs differents sont tres encourageants : une combinaison de plusieurs strategies permet d'atteindre un taux de reconnaissance de 90 %.

https://archivesic.ccsd.cnrs.fr/sic_00001221/document

A. Moises

Papers

Combining model-based and discriminative classifiers: application to handwritten character recognition

Hybrid generative/discriminative classifier for unconstrained character recognition

Self-supervised writer adaptation using perceptive concepts : application to on-line text recognition

Self-supervised adaptation for on-line text recognition

Techniques d'adaptation au scripteur pour la lecture de textes manuscrits dynamiques