M.Tech Final Year Project done by --Paromita Saha
I have done this project in PyCharm Professional Edition. I used OpenCV and Numpy to configure the Object Detection Project.
Full Stucture of the above project "OBJECT DETECTION".
The Network Layer that I used is MOBILENETV3.
The MS-COCO dataset consists 91 classes in total, but only use 82 classes.
The dataset has been trained using SINGLE SHOT DETECTOR (SSD) for localisation of images, before passing through the mobilenet network layers.
To know more about SSD refer this https://towardsdatascience.com/ssd-single-shot-detector-for-object-detection-using-multibox-1818603644ca?gi=8a3bb0e61a08
This above project uses NMS, to understand the topic refer to these images. To know about NMS refer https://www.analyticsvidhya.com/blog/2020/08/selecting-the-right-bounding-box-using-non-max-suppression-with-implementation/
Example of NON_MAXIMUM SUPPRESSION:-
The full SSD-MobileNetV3 model is
Other Object Detectors using Deep Learning for localization of dataset images are
Very easy project for a beginner If you are a new learner in Python, you must try this project.
Thank you all.
NOTE: ALL THE IMAGES HERE ABOVE ARE HAND DRAWN IN MS-WORD. THERE IS NO OTHER COPYRIGHTED IMAGES EXCLUDING THE NUMPY & OPENCV LOGOS.