Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth
World Scientific Annual Review of Artificial Intelligence · 2024
DOI: 10.1142/s2811032324400034 ↗Kurz gefasst
Fachartikel in der Zeitschrift World Scientific Annual Review of Artificial Intelligence, 2024. Verfasst von Chenhang He und 1 weiteren Autor:innen. Die Arbeit nennt 56 Literaturangaben.
Abstract (Originalsprache)
Exploiting geometric features is a common approach to enhance monocular 3D object detection. However, their performance is limited due to the absence of depth information. To address this limitation, an external depth estimator can be employed to predict depth, but this approach significantly reduces the efficiency and flexibility of the model. Instead of relying on a costly depth estimator, we propose a depth-aware monocular 3D object detector that is trained using augmented training data. Specifically, we utilize reference images and their corresponding depth maps to train an efficient rendering module, which synthesizes a variety of photo-realistic images with different virtual depths. By learning from these images, the detector adapts its features to depth variations. Furthermore, we introduce an auxiliary module that guides the network to learn more informative representations from the depth images. Both modules are removed after training, resulting in no additional computational overhead during the final deployment.
Zitieren
APA
He, C., & Zhang, L. (2024). Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth. World Scientific Annual Review of Artificial Intelligence. https://doi.org/10.1142/s2811032324400034
Vancouver
He C, Zhang L. Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth. World Scientific Annual Review of Artificial Intelligence. 2024. doi:10.1142/s2811032324400034
BibTeX
@article{He2024,
author = {He, Chenhang and Zhang, Lei},
title = {Improving Monocular 3D Object Detection by Synthetic Images with Virtual Depth},
journal = {World Scientific Annual Review of Artificial Intelligence},
year = {2024},
doi = {10.1142/s2811032324400034}
}
EUVIMED