🔎👁️Object Detection state-of-the-art👁️🔍

CONCEPT:

The main purpose of this blog is to explain the concept of Object Detecton in a brief and crisp manner. It enables the users to get a better understanding about Object detection and its methodology as the key points are highlighted in the below created blog.

INTRODUCTION :

Object detection is the task of detecting instances of objects of a certain class within an image. This article goes over the most recent state of the art object detectors. The state-of-the-art methods can be categorized into two main types.

# One-stage methods

# Two-stage methods

One-stage methods with example :

• Inference speed ,

• YOLO ,

• SSD

• RetinaNet.

Two-stage methods with example :

• Detection accuracy ,

• Faster R-CNN ,

• Mask R-CNN,

• Cascade R-CNN.

MY FIRST BLOG MADE BY MY OWN SELF MADE VIDEO / IMAGES

SELF MADE VIDEO :

https://drive.google.com/file/d/1yY3gqd3pDXAgxiocNnH2h2IRTtcB6cGB/view?usp=sharing

2D Detectors :

• 2D object detectors give bounding boxes with Four Degrees of Freedom

• This information is crucial to predict important factors like shape , size , position of object

3D Detectors :

• 3D Detectors use data from Camera

• LIDAR or Radar to generate 3D bounding boxes

Person Detection :

• Person detection is a variant of object detection used to detect a primary class "Person" in image or video frames.

• Most modern person detector techniques are trained on frontal and asymmetric views,

Important of Object Detection :

• Object detection is the one of the fundamental problems of computer vision.

  Example :

  Instance segmentation , Image captioning , Object tracking

Image processing techniques :

• In generally don't require historical data for training and are unsupervised in nature.

• OpenCV is a popular tool for image processing tasks

Deep learing methods :

• In generally depend on supervised or unsupervised learning .

• Supervised methods being the standing in computer vision tasks.

IoU :

• IoU metric evaluates the division between the area of overlap and the area of union.

• It evaluates the degree of overlap between ground truth (gt) and predictions (pd).

Specific object detection application :

• Animal detection ,

• Vehicle detection ,

• People counting ,

• Face detecting ,

• Number plate recognition.

Object detection also useful for Data analysis. From the images , we can bring the different types of data.

Based on the image reference we can analysis the data :

Raw data :

DATA ANALYSIS :

Conclusion :

The powerful learning ability and advantages in dealing with blocking , scale transformation and background switches , deep learning based object detection has been a research hotspot in last few years. we propose several promising future directions to gain a through understanding of the object detection landscape. This review is also meaningful for the developments in neural networks systems.

- ❤️D.SUNDARRAJAN❤️