QAL: A Loss for Recall Precision Balance in 3D Reconstruction

TL;DR

QAL is a novel loss function for 3D reconstruction that balances recall and precision, leading to better coverage of thin and under-represented structures across various pipelines and datasets.

Contribution

The paper introduces Quality-Aware Loss (QAL), a new loss function that explicitly decouples recall and precision in 3D reconstruction training objectives.

Findings

QAL improves coverage by +4.3 points over CD and +2.8 over alternatives.

QAL reliably recovers thin structures and under-represented regions.

QAL-trained models show higher grasp scores, indicating better practical utility.

Abstract

Volumetric learning underpins many 3D vision tasks such as completion, reconstruction, and mesh generation, yet training objectives still rely on Chamfer Distance (CD) or Earth Mover's Distance (EMD), which fail to balance recall and precision. We propose Quality-Aware Loss (QAL), a drop-in replacement for CD/EMD that combines a coverage-weighted nearest-neighbor term with an uncovered-ground-truth attraction term, explicitly decoupling recall and precision into tunable components. Across diverse pipelines, QAL achieves consistent coverage gains, improving by an average of +4.3 pts over CD and +2.8 pts over the best alternatives. Though modest in percentage, these improvements reliably recover thin structures and under-represented regions that CD/EMD overlook. Extensive ablations confirm stable performance across hyperparameters and across output resolutions, while full retraining on PCN and ShapeNet demonstrates generalization across datasets and backbones. Moreover, QAL-trained completions yield higher grasp scores under GraspNet evaluation, showing that improved coverage translates directly into more reliable robotic manipulation. QAL thus offers a principled, interpretable, and practical objective for robust 3D vision and safety-critical robotics pipelines