Skip to Main content Skip to Navigation
Conference papers

Un générateur de boites englobantes parcimonieux pour la détection d'objets dans des vidéos

Abstract : The ability to automatically detect object in video stream while dealing with computation time constraint is still not reached by the state of the art algorithms. However, some applications need only the detection of each object at least one time, but not necessarily, the detection of each object in each frame. This context allows to soften algorithmic constraints. Particularly, in this paper, we offer a bounding box proposal algorithm designed to recover each target at least one time in the video with a very small set of boxes per frame. This algorithm achieves interesting result on the VIRAT aerial dataset while handling more than 25 frames per second.
Document type :
Conference papers
Complete list of metadata

Cited literature [12 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01175556
Contributor : Adrien Chan-Hon-Tong Connect in order to contact the contributor
Submitted on : Friday, July 10, 2015 - 4:03:20 PM
Last modification on : Thursday, October 14, 2021 - 9:16:05 AM
Long-term archiving on: : Monday, October 12, 2015 - 11:32:39 AM

File

chanhontong20.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01175556, version 1

Collections

Citation

Adrien Chan-Hon-Tong, Stephane Herbin, Alexandre Boulch. Un générateur de boites englobantes parcimonieux pour la détection d'objets dans des vidéos. XXV colloque Gretsi 2015, Sep 2015, Lyon, France. ⟨hal-01175556⟩

Share

Metrics

Les métriques sont temporairement indisponibles