|
--- |
|
license: apache-2.0 |
|
language: |
|
- en |
|
- tr |
|
metrics: |
|
- accuracy |
|
base_model: |
|
- microsoft/resnet-50 |
|
pipeline_tag: image-classification |
|
library_name: keras |
|
tags: |
|
- image-classification |
|
- resnet50 |
|
- transfer-learning |
|
--- |
|
# ALSATIX ResNet50 Model |
|
|
|
This model is trained to classify images into 5 categories: |
|
|
|
1. **Alkol**: Alcohol-related images |
|
2. **Normal**: Regular images |
|
3. **NSFW**: Not Safe for Work images |
|
4. **Silah**: Weapon-related images |
|
5. **Tutun**: Tobacco-related images |
|
|
|
## Model Architecture |
|
- Base: ResNet50 pre-trained on ImageNet |
|
- Custom top layers: Dense (256 units), Dropout (0.5), Output (5 classes) |
|
|
|
## Usage |
|
|
|
To use this model for image classification: |
|
|
|
```python |
|
from transformers import TFAutoModelForImageClassification, AutoImageProcessor |
|
|
|
model = TFAutoModelForImageClassification.from_pretrained("iammbrn/alsatix_image_control_model") |
|
processor = AutoImageProcessor.from_pretrained("iammbrn/alsatix_image_control_model") |
|
|
|
# Preprocess your image |
|
image = processor(image, return_tensors="pt") |
|
predictions = model(**image) |
