Scribd
Upload
13 views2 pages

Document Comp

The document outlines a project titled 'Image Classification Using Machine Learning' by Shaik Shayaan Ashraf, focusing on using Lobe software to create a custom AI model for fruit classification. The project aims to distinguish between apples, bananas, and oranges, achieving high accuracy through a diverse dataset. Key features of Lobe include no-code model creation, local privacy, and the ability to export models for various applications.

Uploaded by

shaikhshayaan28ashraf
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
13 views2 pages

Document Comp

The document outlines a project titled 'Image Classification Using Machine Learning' by Shaik Shayaan Ashraf, focusing on using Lobe software to create a custom AI model for fruit classification. The project aims to distinguish between apples, bananas, and oranges, achieving high accuracy through a diverse dataset. Key features of Lobe include no-code model creation, local privacy, and the ability to export models for various applications.

Uploaded by

shaikhshayaan28ashraf
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

Page 1: Cover Page

PROJECT TITLE: IMAGE CLASSIFICATION USING MACHINE LEARNING

Submitted by: SHAIKH SHAYAAN ASHRAF

Class: VIII/A

Subject: Computer Science / IT

Roll No: 34

Date: February 11, 2026

Page 2: About Lobe Software


Overview

Lobe is a user-friendly desktop application designed to help people create custom machine
learning models without writing code. It focuses on Image Classification, which is the process
of training a computer to "see" and identify objects in photos.

How it Works

1. Label: Collect and categorize images into different folders.


2. Train: Lobe automatically builds a neural network based on the
provided data.
3. Use: The finished model can be tested using a webcam or by
uploading new images to see if the AI can identify them correctly.

Key Features

 Privacy: All training happens locally on the computer.


 No-Code: No programming knowledge is required.
 Export: Models can be exported to run on websites, mobile apps, or
Raspberry Pi.

SCRATCH

Scratch is a visual, block-based programming language developed by MIT that


simplifies coding by replacing complex text syntax with interlocking graphical blocks.
Users drag and drop color-coded commands—ranging from motion and sound to logic
and variables—onto a workspace to control characters called “sprites” on a stage. This
“jigsaw puzzle” approach prevents common syntax errors, such as missing semicolons,
while teaching fundamental computer science concepts like loops, conditionals, and
event handling.

Would you like me to list the primary color-coded categories and what
each one does?

Page 3: Fruit Classification Project


Objective

To build an AI model that can accurately distinguish between three types of fruits: Apples,
Bananas, and Oranges.

Project Screenshots & Analysis

Observations:

 Dataset: I used approximately 30 images per fruit category.


 Accuracy: As seen in the training sidebar, the model achieved high
confidence levels (e.g., 97%) when the fruit was clearly visible.
 Learning Process: The software highlights "Correct" and "Incorrect"
predictions. If the AI identifies a Banana as an Apple, I can manually
correct it to make the model smarter.

Conclusion

Through this project, I learned that machine learning requires a diverse dataset to be accurate. By
providing images with different angles and lighting, the Lobe AI becomes more reliable at
identifying objects in the real world.

You might also like