17735

On Pre-Trained Image Features and Synthetic Images for Deep Learning

Stefan Hinterstoisser, Vincent Lepetit, Paul Wohlhart, Kurt Konolige
Universite de Bordeaux
arXiv:1710.10710 [cs.CV], (29 Oct 2017)

@article{hinterstoisser2017pretrained,

   title={On Pre-Trained Image Features and Synthetic Images for Deep Learning},

   author={Hinterstoisser, Stefan and Lepetit, Vincent and Wohlhart, Paul and Konolige, Kurt},

   year={2017},

   month={oct},

   archivePrefix={"arXiv"},

   primaryClass={cs.CV}

}

Download Download (PDF)   View View   Source Source   

5319

views

Deep Learning methods usually require huge amounts of training data to perform at their full potential, and often require expensive manual labeling. Using synthetic images is therefore very attractive to train object detectors, as the labeling comes for free, and several approaches have been proposed to combine synthetic and real images for training. In this paper, we show that a simple trick is sufficient to train very effectively modern object detectors with synthetic images only. In order to train them we initialize the part responsible for feature extraction with generic layers pretrained on real data. We show that that freezing these layers and train only the remaining layers with plain OpenGL rendering performs surprisingly well.
Rating: 2.0/5. From 1 vote.
Please wait...

* * *

* * *

HGPU group © 2010-2024 hgpu.org

All rights belong to the respective authors

Contact us: