GroCo: Ground Constraint for Metric Self-Supervised Monocular Depth
This work addresses the challenge of improving metric depth estimation for self-supervised learning in applications like autonomous driving, but it is incremental as it builds on existing ground prior methods.
The paper tackles the problem of metric monocular depth estimation in self-supervised settings, which struggles with scale recovery and generalization across diverse camera poses and datasets. The result shows that their method surpasses existing scale recovery techniques on the KITTI benchmark and enhances generalization, with robust performance across camera rotations and adaptability in zero-shot conditions on datasets like DDAD.
Monocular depth estimation has greatly improved in the recent years but models predicting metric depth still struggle to generalize across diverse camera poses and datasets. While recent supervised methods mitigate this issue by leveraging ground prior information at inference, their adaptability to self-supervised settings is limited due to the additional challenge of scale recovery. Addressing this gap, we propose in this paper a novel constraint on ground areas designed specifically for the self-supervised paradigm. This mechanism not only allows to accurately recover the scale but also ensures coherence between the depth prediction and the ground prior. Experimental results show that our method surpasses existing scale recovery techniques on the KITTI benchmark and significantly enhances model generalization capabilities. This improvement can be observed by its more robust performance across diverse camera rotations and its adaptability in zero-shot conditions with previously unseen driving datasets such as DDAD.