Data Model Details (Fruit Detection)

Detailed description of the data model and Tech used to Build the Application

Images Used
1055 Images of 5 different fruits
Dataset Split
Roboflow API
Link
Project Framework
Sveltekit
About
Fruit detection of apple, pineapple, onion, tomato and watermelon is done using Roboflow API. The model is trained on 1055 images of 5 different fruits. The dataset is split into 93% train, 3% valid and 4% test set. The model is trained for 100 epochs with a batch size of 32. The model is trained on a GPU. The model is deployed using Sveltekit.
Object Detection Algorithm
Region-based Convolutional Neural Networks (R-CNN),Faster R-CNN, YOLO, RetinaNet
Python Library
Open CV, YOLO, NumPy,Scikit-learn