Domain Randomization for Scene-Specific Car Detection and Pose Estimation
Authors:
- Rawal KhirodkarRKRawal Khirodkar
- Donghyun YooDYDonghyun Yoo
- Kris KitaniKKKris Kitani
- 1 January 2019
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1932-1940
Abstract
We address the issue of domain gap when making use of synthetic data to train a scene-specific object detector and pose estimator. While previous works have shown that the constraints of learning a scene-specific model can be leveraged to create geometrically and photometrically consistent synthetic data, care must be taken to design synthetic content which is as close as possible to the real-world data distribution. In this work, we propose to solve domain gap through the use of appearance randomization to generate a wide range of synthetic objects to span the space of realistic images for training. An ablation study of our results is presented to delineate the individual contribution of different components in the randomization process. We evaluate our method on VIRAT, UA-DETRAC, EPFL-Car datasets, where we demonstrate that using scene specific domain randomized synthetic data is better than fine-tuning off-the-shelf models on limited real data.Keywords
All Related Versions
This publication has 18 references indexed in Scilit:
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Visual Domain Adaptation: A survey of recent advancesIEEE Signal Processing Magazine, 2015
- Beyond PASCAL: A benchmark for 3D object detection in the wild2014
- 3D Pose-by-Detection of Vehicles via Discriminatively Reduced Ensembles of Correlation Filters2014
- Vision meets robotics: The KITTI datasetThe International Journal of Robotics Research, 2013
- Teaching 3D geometry to deformable part models2012
- A large-scale benchmark dataset for event recognition in surveillance video2011
- A Survey on Transfer LearningIEEE Transactions on Knowledge and Data Engineering, 2009
- EPnP: An Accurate O(n) Solution to the PnP ProblemInternational Journal of Computer Vision, 2008
- 3D generic object categorization, localization and pose estimation2007