naoto0804/cross-domain-detection
A research implementation of cross-domain weakly-supervised object detection using progressive domain adaptation with SSD-based models.
Not currently ranked — collecting fresh signals.
star history
This repository implements the CVPR 2018 paper on cross-domain object detection through progressive domain adaptation. It uses Chainer and ChainerCV for deep learning-based object detection (SSD300 architecture) and applies domain transfer and pseudo-labeling techniques to adapt models across visual domains (e.g., from real images to watercolor or clipart). The approach enables object detection in target domains using only weak supervision from source domain annotations.
Frequently asked
- What is naoto0804/cross-domain-detection?
- A research implementation of cross-domain weakly-supervised object detection using progressive domain adaptation with SSD-based models.
- Is cross-domain-detection open source?
- Yes — naoto0804/cross-domain-detection is an open-source project tracked on heatdrop.
- What language is cross-domain-detection written in?
- naoto0804/cross-domain-detection is primarily written in Python.
- How popular is cross-domain-detection?
- naoto0804/cross-domain-detection has 437 stars on GitHub.
- Where can I find cross-domain-detection?
- naoto0804/cross-domain-detection is on GitHub at https://github.com/naoto0804/cross-domain-detection.