|
|
--- |
|
|
license: mit |
|
|
tags: |
|
|
- vision |
|
|
- image-classification |
|
|
- flowchart-detection |
|
|
- pytorch |
|
|
- VisionTransformer |
|
|
pipeline_tag: image-classification |
|
|
--- |
|
|
|
|
|
# Flowchart Classification Model |
|
|
|
|
|
## Model Description |
|
|
|
|
|
This is a Vision Transformer (ViT) model fine-tuned for binary classification of images into **flowchart** or **non-flowchart** categories. The model achieves **~85% accuracy** on validation data. |
|
|
|
|
|
- **Model type:** Vision Transformer (ViT-Base) |
|
|
- **Input:** 224x224 RGB images |
|
|
- **Output:** Binary classification (flowchart/non-flowchart) |
|
|
- **Training data:** Custom dataset of 600+ flowchart/non-flowchart images |
|
|
|
|
|
## Intended Uses |
|
|
|
|
|
This model is designed for: |
|
|
|
|
|
- Automated classification of flowchart and non-flowchart image |
|
|
|
|
|
## Github |
|
|
[Link](https://https://github.com/abirmoy/flowchart_classification) |
|
|
|
|
|
## Dataset |
|
|
[Download](https://huggingface.co/abirmoy/flowchart_vit.pth/blob/main/dataset.zip) |
|
|
|
|
|
|
