Dataset Optimization for Real-Time Pedestrian Detection

Dataset Optimization for Real-Time Pedestrian Detection

This paper tackles the problem of data selection for training set generation in the context of near real-time pedestrian detection through the introduction of a training methodology: FairTrain. After highlighting the impact of poorly chosen data on detector performance, we introduce a new data selec...

Full description

Bibliographic Details
Main Authors: Remi Trichet, Francois Bremond
Format: Article
Language:English
Published: IEEE 2018-01-01
Series:IEEE Access
Subjects:
Data Selection
dataset optimization
imbalanced datsets
computer vision
pedestrian detection
real-time application
Online Access:https://ieeexplore.ieee.org/document/8272024/

Internet

https://ieeexplore.ieee.org/document/8272024/

Similar Items