Bayes classification for asynchronous event-based cameras. Fillatre, L. In 2015 23rd European Signal Processing Conference (EUSIPCO), pages 824-828, Aug, 2015. Paper doi abstract bibtex Asynchronous event-based cameras use time encoding to code the pixel intensity values. A time encoding of an input pattern generates a random stream of asynchronous events. An event is defined as a pair containing a timestamp and the variation sign of the input signal since the last emitted event. The goal of this paper is the recognition of the input pattern among a set of several known possibilities from the observation of the event stream. This paper proposes a statistical model of the random event stream based on the physical model of the event-based camera. It also calculates the optimal Bayes classifier which recognizes the input pattern. The numerical complexity of the classifier is rather low. The Bayes risk, which measures the performance of the classifier, is numerically evaluated on simulated data. It is compared to the mean number of events, which entails the power consumption of the camera, exploited to take the decision.
@InProceedings{7362498,
author = {L. Fillatre},
booktitle = {2015 23rd European Signal Processing Conference (EUSIPCO)},
title = {Bayes classification for asynchronous event-based cameras},
year = {2015},
pages = {824-828},
abstract = {Asynchronous event-based cameras use time encoding to code the pixel intensity values. A time encoding of an input pattern generates a random stream of asynchronous events. An event is defined as a pair containing a timestamp and the variation sign of the input signal since the last emitted event. The goal of this paper is the recognition of the input pattern among a set of several known possibilities from the observation of the event stream. This paper proposes a statistical model of the random event stream based on the physical model of the event-based camera. It also calculates the optimal Bayes classifier which recognizes the input pattern. The numerical complexity of the classifier is rather low. The Bayes risk, which measures the performance of the classifier, is numerically evaluated on simulated data. It is compared to the mean number of events, which entails the power consumption of the camera, exploited to take the decision.},
keywords = {cameras;image coding;image recognition;Bayes classification;asynchronous event-based cameras;time encoding;pixel intensity values;asynchronous events;random stream;event-based camera;numerical complexity;optimal Bayes classifier;power consumption;Cameras;Sensors;Encoding;Numerical models;Europe;Signal processing;Neuromorphics;Time encoding;Statistical classification;Event-based camera;Bayes risk},
doi = {10.1109/EUSIPCO.2015.7362498},
issn = {2076-1465},
month = {Aug},
url = {https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104201.pdf},
}
Downloads: 0
{"_id":"GvPTDxxzKcsdXcyuJ","bibbaseid":"fillatre-bayesclassificationforasynchronouseventbasedcameras-2015","authorIDs":[],"author_short":["Fillatre, L."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["L."],"propositions":[],"lastnames":["Fillatre"],"suffixes":[]}],"booktitle":"2015 23rd European Signal Processing Conference (EUSIPCO)","title":"Bayes classification for asynchronous event-based cameras","year":"2015","pages":"824-828","abstract":"Asynchronous event-based cameras use time encoding to code the pixel intensity values. A time encoding of an input pattern generates a random stream of asynchronous events. An event is defined as a pair containing a timestamp and the variation sign of the input signal since the last emitted event. The goal of this paper is the recognition of the input pattern among a set of several known possibilities from the observation of the event stream. This paper proposes a statistical model of the random event stream based on the physical model of the event-based camera. It also calculates the optimal Bayes classifier which recognizes the input pattern. The numerical complexity of the classifier is rather low. The Bayes risk, which measures the performance of the classifier, is numerically evaluated on simulated data. It is compared to the mean number of events, which entails the power consumption of the camera, exploited to take the decision.","keywords":"cameras;image coding;image recognition;Bayes classification;asynchronous event-based cameras;time encoding;pixel intensity values;asynchronous events;random stream;event-based camera;numerical complexity;optimal Bayes classifier;power consumption;Cameras;Sensors;Encoding;Numerical models;Europe;Signal processing;Neuromorphics;Time encoding;Statistical classification;Event-based camera;Bayes risk","doi":"10.1109/EUSIPCO.2015.7362498","issn":"2076-1465","month":"Aug","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104201.pdf","bibtex":"@InProceedings{7362498,\n author = {L. Fillatre},\n booktitle = {2015 23rd European Signal Processing Conference (EUSIPCO)},\n title = {Bayes classification for asynchronous event-based cameras},\n year = {2015},\n pages = {824-828},\n abstract = {Asynchronous event-based cameras use time encoding to code the pixel intensity values. A time encoding of an input pattern generates a random stream of asynchronous events. An event is defined as a pair containing a timestamp and the variation sign of the input signal since the last emitted event. The goal of this paper is the recognition of the input pattern among a set of several known possibilities from the observation of the event stream. This paper proposes a statistical model of the random event stream based on the physical model of the event-based camera. It also calculates the optimal Bayes classifier which recognizes the input pattern. The numerical complexity of the classifier is rather low. The Bayes risk, which measures the performance of the classifier, is numerically evaluated on simulated data. It is compared to the mean number of events, which entails the power consumption of the camera, exploited to take the decision.},\n keywords = {cameras;image coding;image recognition;Bayes classification;asynchronous event-based cameras;time encoding;pixel intensity values;asynchronous events;random stream;event-based camera;numerical complexity;optimal Bayes classifier;power consumption;Cameras;Sensors;Encoding;Numerical models;Europe;Signal processing;Neuromorphics;Time encoding;Statistical classification;Event-based camera;Bayes risk},\n doi = {10.1109/EUSIPCO.2015.7362498},\n issn = {2076-1465},\n month = {Aug},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104201.pdf},\n}\n\n","author_short":["Fillatre, L."],"key":"7362498","id":"7362498","bibbaseid":"fillatre-bayesclassificationforasynchronouseventbasedcameras-2015","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2015/papers/1570104201.pdf"},"keyword":["cameras;image coding;image recognition;Bayes classification;asynchronous event-based cameras;time encoding;pixel intensity values;asynchronous events;random stream;event-based camera;numerical complexity;optimal Bayes classifier;power consumption;Cameras;Sensors;Encoding;Numerical models;Europe;Signal processing;Neuromorphics;Time encoding;Statistical classification;Event-based camera;Bayes risk"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2015url.bib","creationDate":"2021-02-13T17:31:52.362Z","downloads":0,"keywords":["cameras;image coding;image recognition;bayes classification;asynchronous event-based cameras;time encoding;pixel intensity values;asynchronous events;random stream;event-based camera;numerical complexity;optimal bayes classifier;power consumption;cameras;sensors;encoding;numerical models;europe;signal processing;neuromorphics;time encoding;statistical classification;event-based camera;bayes risk"],"search_terms":["bayes","classification","asynchronous","event","based","cameras","fillatre"],"title":"Bayes classification for asynchronous event-based cameras","year":2015,"dataSources":["eov4vbT6mnAiTpKji","knrZsDjSNHWtA9WNT"]}