AutoSynth: Learning to Generate 3D Training Data for Object Point Cloud Registration
This addresses the difficulty of collecting real 3D data for object registration, offering an automated solution for researchers and practitioners in computer vision and robotics.
The paper tackles the problem of generating synthetic 3D training data for point cloud registration by introducing AutoSynth, which automatically curates optimal datasets from millions of possibilities using shape primitives and meta-learning, achieving a 4056.43× speedup with a surrogate network and showing better performance than ModelNet40 on benchmarks like TUD-L and LINEMOD.
In the current deep learning paradigm, the amount and quality of training data are as critical as the network architecture and its training details. However, collecting, processing, and annotating real data at scale is difficult, expensive, and time-consuming, particularly for tasks such as 3D object registration. While synthetic datasets can be created, they require expertise to design and include a limited number of categories. In this paper, we introduce a new approach called AutoSynth, which automatically generates 3D training data for point cloud registration. Specifically, AutoSynth automatically curates an optimal dataset by exploring a search space encompassing millions of potential datasets with diverse 3D shapes at a low cost.To achieve this, we generate synthetic 3D datasets by assembling shape primitives, and develop a meta-learning strategy to search for the best training data for 3D registration on real point clouds. For this search to remain tractable, we replace the point cloud registration network with a much smaller surrogate network, leading to a $4056.43$ times speedup. We demonstrate the generality of our approach by implementing it with two different point cloud registration networks, BPNet and IDAM. Our results on TUD-L, LINEMOD and Occluded-LINEMOD evidence that a neural network trained on our searched dataset yields consistently better performance than the same one trained on the widely used ModelNet40 dataset.