Skip to content

Latest commit

 

History

History
75 lines (56 loc) · 2.39 KB

README.md

File metadata and controls

75 lines (56 loc) · 2.39 KB
Raw

Xtreme-Vision

Build Status License: MIT

Go to PyPI page> Here

This is the Official Repository of Xtreme-Vision. Xtreme-Vision is a High Level Python Library which is built with simplicity in mind for Computer Vision Tasks, such as Object-Detection, Human-Pose-Estimation, Segmentation Tasks, it provides the support of a list of state-of-the-art algorithms, You can Start Detecting with Pretrained Weights as well as You can train the Models On Custom Dataset and with Xtreme-Vision you have the Power to detect/segment only the Objects of your interest

Currently, It Provides the Solution for the following Tasks:

  • Object Detection
  • Pose Estimation
  • Object Segmentation
  • Human Part Segmentation

For Detection with pre-trained models it provides:

  • RetinaNet
  • CenterNet
  • YOLOv4
  • TinyYOLOv4
  • Mask-RCNN
  • DeepLabv3+ (Ade20k)
  • CDCL (Cross Domain Complementary Learning)

For Custom Training It Provides:

  • YOLOv4
  • TinyYOLOv4
  • RetinaNet with (resnet50, resnet101, resnet152)

If You Like this Project, Sponser it here Build Status

Dependencies:

  • tensorflow >= 2.3.0
  • keras
  • opencv-python
  • numpy
  • pillow
  • matplotlib
  • pandas
  • scikit-learn
  • scikit-image
  • imgaug
  • labelme2coco
  • progressbar2
  • scipy
  • h5py
  • configobj

Get Started:

!pip install xtreme-vision

For More Tutorials of Xtreme-Vision, Click Here

YOLOv4 Example

Image Object Detection Using YOLOv4

from xtreme_vision.Detection import Object_Detection

model = Object_Detection()
model.Use_YOLOv4()
model.Detect_From_Image(input_path='kite.jpg',
                        output_path='./output.jpg')

from PIL import Image
Image.open('output.jpg')