Multiple Object Tracking by Efficient Graph Partitioning - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2014

Multiple Object Tracking by Efficient Graph Partitioning


In this paper, we view multiple object tracking as a graphpartitioning problem. Given any object detector, we build the graph ofall detections and aim to partition it into trajectories. To quantifythe similarity of any two detections, we consider local cues such as pointtracks and speed, global cues such as appearance, as well as intermediateones such as trajectory straightness. These different clues are dealt jointlyto make the approach robust to detection mistakes (missing or extradetections). We thus define a Conditional Random Field and optimizeit using an efficient combination of message passing and move-makingalgorithms. Our approach is fast on video batch sizes of hundreds offrames. Competitive and stable results on varied videos demonstrate the robustnessand efficiency of our approach.
Fichier principal
Vignette du fichier
ACCV2014_Tracking.pdf (4.98 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01061450 , version 1 (19-11-2014)


Attribution - NonCommercial


  • HAL Id : hal-01061450 , version 1


Ratnesh Kumar, Guillaume Charpiat, Monique Thonnat. Multiple Object Tracking by Efficient Graph Partitioning. ACCV - 12th Asian Conference on Computer Vision, Michael S. Brown and Tat-Jen Cham and Yasuyuki Matsushita, Nov 2014, Singapore, Singapore. ⟨hal-01061450⟩
284 View
317 Download


Gmail Facebook X LinkedIn More