Computer Vision

This module focuses on two main areas: Human Analysis and Object Detection, obtaining images from a Stereolabs ZED2.

Human Analysis

Several nodes were developed to cover the following main features:

• Face recognition
• Person tracking and re-identification
• Pose and action detection
• Person attributes and description

Object Detection

The main objective for this year was the development of a robust pipeline to make datasets for training YOLO models, focusing on:

• Dataset generation
• Transform object detection datasets to segmentation datasets