MegaSynth: Scaling Up 3D Scene Reconstruction with Synthesized Data
This addresses the data scarcity issue in 3D reconstruction for computer vision researchers, though it is incremental as it builds on existing methods like LRMs with a new dataset.
The paper tackles the problem of scaling up 3D scene reconstruction by training with synthesized data, resulting in improved reconstruction quality by 1.2 to 1.8 dB PSNR across diverse image domains and models trained solely on synthesized data performing comparably to those trained on real data.
We propose scaling up 3D scene reconstruction by training with synthesized data. At the core of our work is MegaSynth, a procedurally generated 3D dataset comprising 700K scenes - over 50 times larger than the prior real dataset DL3DV - dramatically scaling the training data. To enable scalable data generation, our key idea is eliminating semantic information, removing the need to model complex semantic priors such as object affordances and scene composition. Instead, we model scenes with basic spatial structures and geometry primitives, offering scalability. Besides, we control data complexity to facilitate training while loosely aligning it with real-world data distribution to benefit real-world generalization. We explore training LRMs with both MegaSynth and available real data. Experiment results show that joint training or pre-training with MegaSynth improves reconstruction quality by 1.2 to 1.8 dB PSNR across diverse image domains. Moreover, models trained solely on MegaSynth perform comparably to those trained on real data, underscoring the low-level nature of 3D reconstruction. Additionally, we provide an in-depth analysis of MegaSynth's properties for enhancing model capability, training stability, and generalization, as well as application to other tasks.