Keypoint detection is an essential component for the object registration and alignment. In this work, we reckon keypoint detection as information compression, and force the model to distill out irrelevant points of an object. Based on this, we propose UKPGAN, a general self-supervised 3D keypoint detector where keypoints are detected so that they could reconstruct the original object shape. Two modules: GAN-based keypoint sparsity control and salient information distillation modules are proposed to locate those important keypoints. Extensive experiments show that our keypoints align well with human annotated keypoint labels, and can be applied to SMPL human bodies under various non-rigid deformations. Furthermore, our keypoint detector trained on clean object collections generalizes well to real-world scenarios, thus further improves geometric registration when combined with off-the-shelf point descriptors. Repeatability experiments show that our model is stable under both rigid and non-rigid transformations, with local reference frame estimation. Our code is available on https://github.com/qq456cvb/UKPGAN.
翻译:关键点检测是物体登记和校正的一个基本组成部分。 在此工作中, 我们将关键点检测视为信息压缩, 并强制该模型提取一个对象的不相干点 。 基于此, 我们提议使用UKPGAN, 一个通用的自我监督的三维关键点检测器, 以便检测到关键点, 从而可以重建原始对象形状 。 两个模块 : 基于 GAN 的 关键点宽度控制和突出信息蒸馏模块, 以定位这些重要的关键点 。 广泛的实验显示, 我们的关键点与人类的注释关键点标签非常吻合, 并且可以在各种非硬质的变形下应用到 SMPL 人体中。 此外, 我们接受过清洁目标收藏训练的关键点检测器, 能够与现实世界情景相匹配, 从而进一步改善与外点描述器相结合的几何测量登记 。 重复性实验显示, 我们的模型在硬和非硬质的变换中保持稳定, 并附有本地的参考框架估计 。 我们的代码可以在 https://github.com/ qq456cv/ / UKPGANANAN 。