Long-tail Internet photo reconstruction

TL;DR

This paper introduces MegaDepth-X, a dataset and sampling strategy to improve 3D reconstruction of long-tail, sparse, and noisy internet images, advancing 3D foundation models.

Contribution

It presents MegaDepth-X and a sampling method to enhance 3D reconstruction robustness in long-tail internet photo collections.

Findings

Finetuning with MegaDepth-X improves reconstruction in sparse scenes.

Enhanced models handle symmetric and repetitive scenes better.

Method maintains performance on standard dense 3D benchmarks.

Abstract

Internet photo collections exhibit an extremely long-tailed distribution: a few famous landmarks are densely photographed and easily reconstructed in 3D, while most real-world sites are represented with sparse, noisy, uneven imagery beyond the capabilities of both classical and learned 3D methods. We believe that tackling this long-tail regime represents one of the next frontiers for 3D foundation models. Although reliable ground-truth 3D supervision from sparse scenes is challenging to acquire, we observe that it can be effectively simulated by sampling sparse subsets from well-reconstructed Internet landmarks. To this end, we introduce MegaDepth-X, a large dataset of 3D reconstructions with clean, dense depth, together with a strategy for sampling sets of training images that mimic camera distributions in long-tail scenes. Finetuning 3D foundation models with these components yields robust reconstructions under extreme sparsity, and also enables more reliable reconstruction in symmetric and repetitive scenes, while preserving generalization to standard, dense 3D benchmark datasets.