pointrcnn运行_PointCNN代码阅读笔记

Readme阅读

代码文件内容

X-conv及PointCNN架构：pointcnn.py

分类任务超参数：pointcnn_cls.py

分割任务相关：pointcnn_seg.py

关于 X-Conv 和 X-DeConv 的参数

以shapenet_x8_2048_fps.py 文件的代码为例

xconv_param_name = ('K', 'D', 'P', 'C', 'links')

xconv_params = [dict(zip(xconv_param_name, xconv_param)) for xconv_param in

[(8, 1, -1, 32 * x, []),

(12, 2, 768, 32 * x, []),

(16, 2, 384, 64 * x, []),

(16, 6, 128, 128 * x, [])]]

xdconv_param_name = ('K', 'D', 'pts_layer_idx', 'qrs_layer_idx')

xdconv_params = [dict(zip(xdconv_param_name, xdconv_param)) for xdconv_param in

[(16, 6, 3, 2),

(12, 6, 2, 1),

(8, 6, 1, 0),

(8, 4, 0, 0)]]

xconv_params

K：近邻数

D：膨胀率

P：输出代表点的数量(-1表示输出输入中的所有点)<