Saved in:
Bibliographic Details
Main Authors: Luginov, Albert, Shahzad, Muhammad
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.14177
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910576403283968
author Luginov, Albert
Shahzad, Muhammad
author_facet Luginov, Albert
Shahzad, Muhammad
contents 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 acknowledgments are available at https://github.com/xapaxca/nimbled .
format Preprint
id arxiv_https___arxiv_org_abs_2408_14177
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle NimbleD: Enhancing Self-supervised Monocular Depth Estimation with Pseudo-labels and Large-scale Video Pre-training
Luginov, Albert
Shahzad, Muhammad
Computer Vision and Pattern Recognition
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 acknowledgments are available at https://github.com/xapaxca/nimbled .
title NimbleD: Enhancing Self-supervised Monocular Depth Estimation with Pseudo-labels and Large-scale Video Pre-training
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2408.14177