目录
3.1 Multi-scale Feature Extraction
3.2 Learning-based Patchmatch(基于学习的Patchmatch)
3.2.1 Initialization and Local Perturbation(初始化/局部扰动)
3.2.2 Adaptive Propagation(自适应传播)
3.2.3 Adaptive Evaluation (适应性评估)
本文推荐读完代码再看!!!!!!!!!
1.摘要部分
We present PatchmatchNet, a novel and learnable cascade formulation of Patchmatch for high-resolution multi-view stereo. With high computation speed and low memory requirement, PatchmatchNet can process higher resolution imagery and is more suited to run on resource limited devices than competitors that employ 3D cost volume regularization. For the first time we introduce an iterative multi-scale Patchmatch in an end-to-end trainable architecture and improve the Patchmatch core algorithm with a novel and learned adaptive propagation and evaluation scheme for each iteration.