search

Vegeta2020 / SE-SSD

SE-SSD: Self-Ensembling Single-Stage Object Detector From Point Cloud, CVPR 2021.
Apache License 2.0
811 stars 128 forks source link

Why use intensity transform in sa_da #103

Open xuwentai opened 11 months ago

xuwentai commented 11 months ago

Thank you for your excellent work.I am interested in sa Da is quite interested and would like to study it carefully. I am interested in sa Da_ The 180 line code in the v2. py file is a bit confusing.

180                        # for intensity transform
181                        new_to_swap_points_intensity = recover_points_intensity_by_ratio(
182                            swapped_points_intensity_ratio, to_swap_points[:, -1:].max(),
183                            to_swap_points[:, -1:].min())
184                        new_swapped_points_intensity = recover_points_intensity_by_ratio(
185                            to_swap_points_intensity_ratio, swapped_points[:, -1:].max(),
186                            swapped_points[:, -1:].min())

Use the last column to_swap_points of as the strength value, but it only contains xyz. The shape of the input points is points: [num_points, 3], and it does not contain the strength value, Can the author provide a brief explanation，thankyou.