The field of computer vision is witnessing significant advancements in 3D object detection and crowd analysis. Recent developments have focused on improving the accuracy and efficiency of 3D object detection models, particularly in scenarios involving point clouds and crowded scenes. Researchers have proposed innovative solutions, such as slice-based representations and dual-stream graph convolutional networks, to address the challenges posed by complex environments. Furthermore, there is a growing interest in developing models that can effectively estimate crowd density and detect objects in crowded scenes, with applications in areas like autonomous driving and surveillance. Notable papers in this area include PointSlice, which achieves high detection accuracy and inference speed, and CrowdQuery, which introduces a density-guided query module for enhanced 2D and 3D detection in crowded scenes. Overall, these advancements have the potential to significantly impact various applications and pave the way for future research in this field.
Advancements in 3D Object Detection and Crowd Analysis
Sources
PointSlice: Accurate and Efficient Slice-Based Representation for 3D Object Detection from Point Clouds
Decoupling Bidirectional Geometric Representations of 4D cost volume with 2D convolution
DSGC-Net: A Dual-Stream Graph Convolutional Network for Crowd Counting via Feature Correlation Mining
Count2Density: Crowd Density Estimation without Location-level Annotations
PI3DETR: Parametric Instance Detection of 3D Point Cloud Edges with a Geometry-Aware 3DETR
FlowSeek: Optical Flow Made Easier with Depth Foundation Models and Motion Bases
3DPillars: Pillar-based two-stage 3D object detection
StripDet: Strip Attention-Based Lightweight 3D Object Detection from Point Cloud
MIORe & VAR-MIORe: Benchmarks to Push the Boundaries of Restoration
CrowdQuery: Density-Guided Query Module for Enhanced 2D and 3D Detection in Crowded Scenes