Max Robotics — Global Service Robot Directory + US Certification

About

Xiaoke Shen is a computer vision researcher specializing in 3D object detection, instance segmentation, and multimodal deep learning. His work sits at the intersection of RGB and depth-based perception, tackling one of the field's most pressing challenges: accurately localizing and identifying objects in three-dimensional space using heterogeneous sensor data. Shen's most influential contribution, *Frustum VoxNet* (2020), introduced a novel pipeline for 3D object detection from RGB-D or depth-only point clouds, garnering 37 citations and establishing him as a notable voice in the volumetric detection community. This work elegantly bridges 2D object detection with 3D spatial reasoning using frustum-based voxel representations. His 2019 survey on 2D/3D object classification and detection (19 citations) further demonstrates his breadth, providing the research community with a structured overview of deep learning approaches across modalities. More recently, Shen has explored cross-modality transfer learning through *simCrossTrans* (2022), proposing methods to leverage knowledge across RGB, depth, and other domains using both ConvNets and Vision Transformers — reflecting his adaptability to emerging architectures. Collectively, his work addresses critical needs in robotics, autonomous systems, and location-sensitive applications where precise 3D scene understanding is essential.