MeSa: Masked, Geometric, and Supervised Pre-training for Monocular Depth Estimation

6 Oct 2023  ·  Muhammad Osama Khan, Junbang Liang, Chun-Kai Wang, Shan Yang, Yu Lou ·

Pre-training has been an important ingredient in developing strong monocular depth estimation models in recent years. For instance, self-supervised learning (SSL) is particularly effective by alleviating the need for large datasets with dense ground-truth depth maps. However, despite these improvements, our study reveals that the later layers of the SOTA SSL method are actually suboptimal. By examining the layer-wise representations, we demonstrate significant changes in these later layers during fine-tuning, indicating the ineffectiveness of their pre-trained features for depth estimation. To address these limitations, we propose MeSa, a comprehensive framework that leverages the complementary strengths of masked, geometric, and supervised pre-training. Hence, MeSa benefits from not only general-purpose representations learnt via masked pre training but also specialized depth-specific features acquired via geometric and supervised pre-training. Our CKA layer-wise analysis confirms that our pre-training strategy indeed produces improved representations for the later layers, overcoming the drawbacks of the SOTA SSL method. Furthermore, via experiments on the NYUv2 and IBims-1 datasets, we demonstrate that these enhanced representations translate to performance improvements in both the in-distribution and out-of-distribution settings. We also investigate the influence of the pre-training dataset and demonstrate the efficacy of pre-training on LSUN, which yields significantly better pre-trained representations. Overall, our approach surpasses the masked pre-training SSL method by a substantial margin of 17.1% on the RMSE. Moreover, even without utilizing any recently proposed techniques, MeSa also outperforms the most recent methods and establishes a new state-of-the-art for monocular depth estimation on the challenging NYUv2 dataset.

PDF Abstract

Datasets


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Monocular Depth Estimation NYU-Depth V2 MeSa RMSE 0.238 # 6
absolute relative error 0.066 # 6
Delta < 1.25 0.964 # 6
Delta < 1.25^2 0.995 # 7
Delta < 1.25^3 0.999 # 3
log 10 0.029 # 6

Methods


No methods listed for this paper. Add relevant methods here