Xin Wang
Research Subjects
Xin Wang has a total of 6 co-authors. Here are the top co-authors by number of shared publications:
- Xin Yang
National University of Singapore
2 shared publications - Yuan Yuan
Huawei Technologies Ltd
2 shared publications Michael Bi Mi
Huawei International Pte Ltd
1 shared publication- Michael Bi Mi
Huawei Technologies Co., Ltd
1 shared publication Robby T. Tan
National University of Singapore Huawei International Pte Ltd Yale-NUS College
1 shared publication- Robby T. Tan
National University of Singapore
1 shared publication
Latest Publications
- 26 February 2023
- Published by Springer Nature
- p. 303-318
Object Detection in Foggy Scenes by Embedding Depth and Reconstruction into Domain Adaptation
- 26 February 2023
- Published by Springer Nature
- p. 303-318
Object Detection in Foggy Scenes by Embedding Depth and Reconstruction into Domain Adaptation
- Open Access
Object Detection in Foggy Scenes by Embedding Depth and Reconstruction into Domain Adaptation
Most existing domain adaptation (DA) methods align the features based on the domain feature distributions and ignore aspects related to fog, background and target objects, rendering suboptimal performance. In our DA framework, we retain the depth and background information during the domain feature alignment. A consistency loss between the generated depth and fog transmission map is introduced to strengthen the retention of the depth information in the aligned features. To address false object features potentially generated during the DA process, we propose an encoder-decoder framework to reconstruct the fog-free background image. This reconstruction loss also reinforces the encoder, i.e., our DA backbone, to minimize false object features.Moreover, we involve our target data in training both our DA module and our detection module in a semi-supervised manner, so that our detection module is also exposed to the unlabeled target data, the type of data used in the testing stage. Using these ideas, our method significantly outperforms the state-of-the-art method (47.6 mAP against the 44.3 mAP on the Foggy Cityscapes dataset), and obtains the best performance on multiple real-image public datasets. Code is available at: https://github.com/VIML-CVDL/Object-Detection-in-Foggy-Scenes
Open AccessObject Detection in Foggy Scenes by Embedding Depth and Reconstruction into Domain Adaptation
Most existing domain adaptation (DA) methods align the features based on the domain feature distributions and ignore aspects related to fog, background and target objects, rendering suboptimal performance. In our DA framework, we retain the depth and background information during the domain feature alignment. A consistency loss between the generated depth and fog transmission map is introduced to strengthen the retention of the depth information in the aligned features. To address false object features potentially generated during the DA process, we propose an encoder-decoder framework to reconstruct the fog-free background image. This reconstruction loss also reinforces the encoder, i.e., our DA backbone, to minimize false object features.Moreover, we involve our target data in training both our DA module and our detection module in a semi-supervised manner, so that our detection module is also exposed to the unlabeled target data, the type of data used in the testing stage. Using these ideas, our method significantly outperforms the state-of-the-art method (47.6 mAP against the 44.3 mAP on the Foggy Cityscapes dataset), and obtains the best performance on multiple real-image public datasets. Code is available at: https://github.com/VIML-CVDL/Object-Detection-in-Foggy-Scenes