https://github.com/ashwaniydv/zero-shot-image-recognition

IIT Patna Academic Project | CS244-Introduction to data science | Image classification: Zero-shot Image Recognition
https://github.com/ashwaniydv/zero-shot-image-recognition

object-detection transfer-learning word2vec zero-shot zero-shot-classification zero-shot-learning

Last synced: 5 months ago
JSON representation

IIT Patna Academic Project | CS244-Introduction to data science | Image classification: Zero-shot Image Recognition

• Host: GitHub
• URL: https://github.com/ashwaniydv/zero-shot-image-recognition
• Owner: ashwaniYDV
• Created: 2020-06-13T13:00:31.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2024-01-25T17:08:38.000Z (over 1 year ago)
• Last Synced: 2025-04-02T21:42:37.644Z (6 months ago)
• Topics: object-detection, transfer-learning, word2vec, zero-shot, zero-shot-classification, zero-shot-learning
• Language: Python
• Homepage:
• Size: 92.8 MB
• Stars: 5
• Watchers: 1
• Forks: 1
• Open Issues: 26
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Zero-shot Image Recognition: Image classification

### Academic Project (CS244-Introduction to data science)

## zero-shot-learning

Implementation of Zero-Shot Learning algorithm  

  

Zero-Shot learning method aims to solve a task without receiving any example of that task at training phase.  

It simply allows us to recognize objects *we have not seen before*.  

  

### Classes  

**Train Classes:**  

arm, boy, bread, chicken, child, computer, ear, house, leg, sandwich, television, truck, vehicle, watch, woman  

**Zero-Shot Classes:**  

car, food, hand, man, neck  

  

### Usage  

$ **python3**  detect_object.py  input-image-path  

  

### Example  

$ **cd**  src  

$ **python3**  detect_object.py  ../test.jpg

--- Top-5 Prediction ---  

1- vehicle  

2- truck  

3- car  

4- house  

5- watch  

  

![Example Image](test.jpg)

*Test image is a beautiful Jaguar.*  

*All related prediction results are ranked in first three.* 

  

P.S. Remember, the prediction results are only allowed to be among above classes (train and zero-shot classes).  

Algorithm will fail (although it will do its best to predict most related class) in case you try to detect an object from different other classes.

### Model Summary 

![Example Image](model.png)

### ZSL Model Summary

![Example Image](zslmodel.png)