File size: 1,181 Bytes
4e92234
255515f
4e92234
 
 
 
 
 
 
 
 
f2c2d5d
 
 
 
 
8109504
 
f2c2d5d
8109504
 
 
 
 
 
 
 
 
 
 
f2c2d5d
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
---
title: Image Classification - Biodegradable vs Nonbiodegradable
emoji: 📚
colorFrom: pink
colorTo: indigo
sdk: gradio
sdk_version: 2.8.10
app_file: app.py
pinned: false
---

This app was built using 2 machine learning models:
- CLIP - A neural network introduced by OpenAI - see https://openai.com/blog/clip
- Logistic regression

These 2 models where pre trained in the following ways:
- CLIP - pre trained variety of images with a wide variety of natural language supervision that’s available on the internet.
- Logistic regression classifier - trained on a transformed data set which was originated from biodegradable and non biodegradable available on Kaggle -   https://www.kaggle.com/rayhanzamzamy/non-and-biodegradable-waste-dataset

The Logistic regression classifier can be used by the pickle file - lr-model.pickel.

The steps of the calculations are as follows:

a. Get an image from the user.

b. Get features of the image by using CLIP neural network.

c. Use pre trained logistic regression classifier to predict the image class.

d. Display class name.