1. Introduction

Lane and object detection (Yolo V5 & openCV)

Real-time person detection in manufacturing production lines.

Applied AI system based on the YOLOv7 algorithm trained for aircraft detection

Read more at: https://viso.ai/deep-learning/object-detection/

1.2 Data sources

1.2.1 Image ⇒ 3D Bounding Box

YOLO3D: A 3D Object Detection Model (source)

1.2.2 Point Cloud ⇒ 3D Bounding Box

3D Cascade R-CNN: LiDAR Object Detection (source)

1.3 Strategies

1. improve the accuracy of object location
   1. more informative and robust representation
   2. coarse-to-fine Localization ⇒ feature alignment
   3. attention \ transformer ⇒ feature alignment

2 Categories