name: "TF Keras CIFAR-10 Model" framework: "TensorFlow" description: "A convolutional neural network trained on CIFAR-10 dataset" tags:
- Classification
- CIFAR-10
- Convolutional Neural Network inputs:
- "32 x 32 x 3 RGB images" architecture:
- "Flatten Layer"
- "Dense Layer with 128 neurons and a Leaky ReLU Activation"
- "Dense Layer with 128 neurons and a Leaky ReLU Activation"
- "Dense Layer with 10 neurons and a Sigmoid Activation" metrics:
- "Optimizer: Adam"
- "Loss Function: Sparse Categorical Crossentropy"
- "Accuracy: 96.75%" datasets:
- "CIFAR-10" outputs:
- "10 labels" sample_usages:
- "Classification of airborne and land-based images"