Soft Masked Transformer for Point Cloud Processing with Skip Attention-Based Upsampling

arXiv:2403.14124v1 Announce Type: new
Abstract: Point cloud processing methods leverage local and global point features %at the feature level to cater to downstream tasks, yet they often overlook the task-level context inherent in point clouds during the encoding stage. We argue that integrating task-level information into the encoding stage significantly enhances performance. To that end, we propose SMTransformer which incorporates task-level information into a vector-based transformer by utilizing a soft mask generated from task-level queries and keys to learn the …

abstract arxiv attention cloud context cs.cv encoding feature features global information processing stage tasks transformer type