Explaining Human Preferences via Metrics for Structured 3D Reconstruction
This work addresses the need for reliable evaluation metrics in 3D reconstruction, which is crucial for researchers and practitioners in computer vision and graphics, though it is incremental in nature.
The paper tackles the problem of evaluating structured 3D reconstructions by analyzing automated metrics, proposing unit tests, and providing context-aware recommendations, with a result of introducing a learned metric distilled from human expert judgments.
"What cannot be measured cannot be improved" while likely never uttered by Lord Kelvin, summarizes effectively the driving force behind this work. This paper presents a detailed discussion of automated metrics for evaluating structured 3D reconstructions. Pitfalls of each metric are discussed, and an analysis through the lens of expert 3D modelers' preferences is presented. A set of systematic "unit tests" are proposed to empirically verify desirable properties, and context aware recommendations regarding which metric to use depending on application are provided. Finally, a learned metric distilled from human expert judgments is proposed and analyzed. The source code is available at https://github.com/s23dr/wireframe-metrics-iccv2025