Object Detection and Tracking

Object detection is a computer technology related to computer vision and image processing that deals with detecting instances of semantic objects of a certain class (such as humans, buildings, or cars) in digital images and videos.

Environment

I have tested on Ubuntu 16.04/18.04. The code may work on other systems.

[Ubuntu-Deep-Learning-Environment-Setup]

• Ubuntu 16.04 / 18.04

• ROS Kinetic / Melodic

• GTX 1080Ti / RTX 2080Ti

• python 2.7 / 3.6

Installation

Clone the repository

git clone https://github.com/yehengchen/Object-Detection-and-Tracking.git

[OneStage]

YOLO: Real-Time Object Detection and Tracking

• How to train a YOLO model on custom images: YOLOv3 - [Link] / YOLOv4 - [Link]

• YOLOv4 + Deep_SORT - Pedestrian Counting & Social Distance - [Link]

• YOLOv3 + Deep_SORT - Pedestrian&Car Counting - [Link]

• YOLOv3 + SORT - Pedestrian Counting - [Link]

Darknet_ROS: Real-Time Object Detection and Grasp Detection With ROS

• YOLOv3 + ROS Kinetic - For small Custom Data - [Link]

• YOLOv3 + ROS Melodic - Robot Grasp Detection - [Link]

• Parts-Arrangement-Robot - [Link]

• YOLOv3 + OpenCV + ROS Melodic - Object Detection (Rotated) - [Link]

SSD: Single Shot MultiBox Detector

• How to train a SSD model on own images - [Link]

[TwoStage]

R-CNN: Region-based methods

Fast R-CNN / Faster R-CNN / Mask R-CNN

How to train a Mask R-CNN model on own images - [Link]

• Mask R-CNN + ROS Kinetic - [Link]

This project is ROS package of Mask R-CNN algorithm for object detection and segmentation.

COCO & VOC Datasets

• COCO dataset and Pascal VOC dataset - [Link]

• How to get it working on the COCO dataset coco2voc - [Link]

• Convert Dataset2Yolo - COCO / VOC - [Link]

CV & Robotics Paper List (3D object detection & 6D pose estimation) - [Link]

PapersWithCode: Browse > Computer Vision > Object Detection - [Link]

ObjectDetection Two-stage vs One-stage Detectors - [Link]

ObjectDetection mAP & IoU - [Link]