---
tags:
- onnx
- image-classification
- cifar10
- dropout
- aidge
pipeline_tag: image-classification
datasets:
- cifar10
metrics:
- accuracy
model-index:
- name: Custom ResNet-18 with integrated Dropout layers
  results:
  - task:
      type: image-classification
      name: Image Classification
    dataset:
      name: CIFAR-10
      type: cifar10
    metrics:
    - type: accuracy
      value: 83.96%
language:
- en
base_model:
- resnet-18
---

# Custom ResNet-18 with integrated Dropout Layers

This is a **custom ResNet-18** ONNX model implemented in **PyTorch** with integrated **Dropout** layers.
It was trained on the **CIFAR-10** dataset for image classification tasks. 
The model has been exported using **opset version 15** and is fully compatible with the **Aidge** platform.

## Details

- **Architecture**: Customized ResNet-18 with integrated Dropout layers
- **Trained on**: CIFAR-10 (60,000 32x32 color images, 10 classes)
- **Image Preprocessing**: Images were resized to `128×128`
- **Data Normalization**: `mean = [0.4914, 0.4822, 0.4465]` ; `std = [0.2023, 0.1994, 0.2010]`
- **Dropout Probability**: 0.3
- **ONNX opset version**: 15
- **Conversion tool**: PyTorch → ONNX

---