NimbleD: Enhancing Self-Supervised Monocular Depth Estimation with Pseudo-Labels and Large-Scale Video Pre-Training
Abstract
We introduce NimbleD, an efficient self-supervised monocular depth estimation learning framework that incorporates supervision from pseudo-labels generated by a large vision model. This framework does not require camera intrinsics, enabling large-scale pre-training on publicly available videos. Our straightforward yet effective learning strategy significantly enhances the performance of fast and lightweight models without introducing any overhead, allowing them to achieve performance comparable to state-of-the-art self-supervised monocular depth estimation models. This advancement is particularly beneficial for virtual and augmented reality applications requiring low latency inference. The source code, model weights, and attributions are available at https://github.com/xapaxca/nimbled .
Cite
Text
Luginov and Shahzad. "NimbleD: Enhancing Self-Supervised Monocular Depth Estimation with Pseudo-Labels and Large-Scale Video Pre-Training." European Conference on Computer Vision Workshops, 2024. doi:10.1007/978-3-031-92387-6_18Markdown
[Luginov and Shahzad. "NimbleD: Enhancing Self-Supervised Monocular Depth Estimation with Pseudo-Labels and Large-Scale Video Pre-Training." European Conference on Computer Vision Workshops, 2024.](https://mlanthology.org/eccvw/2024/luginov2024eccvw-nimbled/) doi:10.1007/978-3-031-92387-6_18BibTeX
@inproceedings{luginov2024eccvw-nimbled,
title = {{NimbleD: Enhancing Self-Supervised Monocular Depth Estimation with Pseudo-Labels and Large-Scale Video Pre-Training}},
author = {Luginov, Albert and Shahzad, Muhammad},
booktitle = {European Conference on Computer Vision Workshops},
year = {2024},
pages = {235-251},
doi = {10.1007/978-3-031-92387-6_18},
url = {https://mlanthology.org/eccvw/2024/luginov2024eccvw-nimbled/}
}