Meta-Optimization for Higher Model Generalizability in Single-Image Depth Prediction
This work addresses robustness for indoor depth prediction in practical applications, though it is incremental as it applies meta-learning to a new domain.
The paper tackles the problem of low model generalizability in single-image depth prediction for unseen indoor datasets by using gradient-based meta-learning to improve zero-shot cross-dataset inference, resulting in up to a 29.4% improvement in prior performance and consistent gains in accuracy and robustness over baselines.
Model generalizability to unseen datasets, concerned with in-the-wild robustness, is less studied for indoor single-image depth prediction. We leverage gradient-based meta-learning for higher generalizability on zero-shot cross-dataset inference. Unlike the most-studied image classification in meta-learning, depth is pixel-level continuous range values, and mappings from each image to depth vary widely across environments. Thus no explicit task boundaries exist. We instead propose fine-grained task that treats each RGB-D pair as a task in our meta-optimization. We first show meta-learning on limited data induces much better prior (max +29.4\%). Using meta-learned weights as initialization for following supervised learning, without involving extra data or information, it consistently outperforms baselines without the method. Compared to most indoor-depth methods that only train/ test on a single dataset, we propose zero-shot cross-dataset protocols, closely evaluate robustness, and show consistently higher generalizability and accuracy by our meta-initialization. The work at the intersection of depth and meta-learning potentially drives both research streams to step closer to practical use.