Skip to content

A Tensorflow2.0 implementation of Single Shot Detector

Notifications You must be signed in to change notification settings

isabella232/SSD-Tensorflow

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

29 Commits
 
 
 
 
 
 

Repository files navigation

Single Shot Detector for Densely Packed Objects

  • A TensorFlow implementation of object detection paper: SSD - Single Shot MultiBox Detector.
  • Replaced the existing VGG backbone with DenseNet to achieve better performance.
  • Calibrated the default anchors to better suit for the new dataset
  • Trained the model on the SKU110K dataset consisting of 12k grocery store images with each image having 200 objects on average, often appearing similar or identical and positioned in close proximity.

Model

Result

Matched default boxes

Pretrained Weights Download

@misc{liu2015ssd,
    title={SSD: Single Shot MultiBox Detector},
    author={Wei Liu and Dragomir Anguelov and Dumitru Erhan and Christian Szegedy and Scott Reed and Cheng-Yang Fu and Alexander                       C.Berg},
    year={2015},
    eprint={1512.02325},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

About

A Tensorflow2.0 implementation of Single Shot Detector

Resources

Code of conduct

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 100.0%