Skip to content

This repository contains the python version of the source code for the experiments carried out for the On-line Object Detection and Instance Segmentation project.

hsp-iit/online-detection

Repository files navigation

On-line Object Detection and Instance Segmentation

This repository contains the python version of the source code for the experiments carried out for the On-line Object Detection and Instance Segmentation project.

Abstract

Object detection and instance segmentation are fundamental tasks for robots interacting within an environment. While stunningly effective, state-of-the-art deep learning methods require huge amounts of labeled images and long training sessions which does not favour such scenarios. In this project, we aim at designing algorithmic solutions to alleviate these requirements for this task, while preserving the state-of-the-art precision and reliability. The proposed methods are typically validated on both computer vision and robotics datasets. This repository allows to reproduce the main experiments of the proposed works and allows the user to test the pipeline with other datasets.

Description

pipeline_orpn_oos_reviewed_2_compressed

This picture presents the current architecture of the pipeline. The Feature Extraction Module relies on the Mask R-CNN architecture and the proposed On-line RPN, to extract deep features and predict RoIs from each input image. The On-line Detection Module performs RoIs classification and refinement, providing as output the detections for the input image. The green blocks are trained off-line on the FEATURE-TASK, while the yellow blocks are trained on-line on the TARGET-TASK.

Installation guide

You can find the instructions for installation at this link.

Experiments

We provide the links to instructions to reproduce the main experiments of the presented works.

Learn Fast, Segment Well: Fast Object Segmentation Learning on the iCub Robot

You can find the instructions to replicate experiments at this link.

Fast Object Segmentation Learning with Kernel-based Methods for Robotics

You can find the instructions to replicate experiments at this link.

Fast Region Proposal Learning for Object Detection for Robotics

You can find the instructions to replicate experiments at this link.

References

If you use this code, please, cite the following works:

@ARTICLE{ceola2022tro,
  author={Ceola, Federico and Maiettini, Elisa and Pasquale, Giulia and Meanti, Giacomo and Rosasco, Lorenzo and Natale, Lorenzo},
  journal={IEEE Transactions on Robotics}, 
  title={Learn Fast, Segment Well: Fast Object Segmentation Learning on the iCub Robot}, 
  year={2022},
  volume={38},
  number={5},
  pages={3154-3172},
  doi={10.1109/TRO.2022.3164331}}
@INPROCEEDINGS{ceola2021oos,
  author={Ceola, Federico and Maiettini, Elisa and Pasquale, Giulia and Rosasco, Lorenzo and Natale, Lorenzo},
  booktitle={2021 IEEE International Conference on Robotics and Automation (ICRA)}, 
  title={Fast Object Segmentation Learning with Kernel-based Methods for Robotics}, 
  year={2021},
  volume={},
  number={},
  pages={13581-13588},
  doi={10.1109/ICRA48506.2021.9561758}
}
@Article{maiettini2019a,
	author="Maiettini, Elisa
	and Pasquale, Giulia
	and Rosasco, Lorenzo
	and Natale, Lorenzo",
	title="On-line object detection: a robotics challenge",
	journal="Autonomous Robots",
	year="2019",
	month="Nov",
	day="25",
	issn="1573-7527",
	doi="10.1007/s10514-019-09894-9",
	url="https://doi.org/10.1007/s10514-019-09894-9"
}
@INPROCEEDINGS{maiettini2018,
	author={E. Maiettini and G. Pasquale and L. Rosasco and L. Natale},
	booktitle={2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
	title={Speeding-up Object Detection Training for Robotics with FALKON},
	year={2018},
	month={Oct},
}
@article{ceola2020rpn,
  title={Fast region proposal learning for object detection for robotics},
  author={Ceola, Federico and Maiettini, Elisa and Pasquale, Giulia and Rosasco, Lorenzo and Natale, Lorenzo},
  journal={arXiv preprint arXiv:2011.12790},
  year={2020}
}

About

This repository contains the python version of the source code for the experiments carried out for the On-line Object Detection and Instance Segmentation project.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages