| | --- |
| | license: apache-2.0 |
| | --- |
| | # Anime Face Generator - DCGAN 🎨✨ |
| |
|
| | This is a **Deep Convolutional Generative Adversarial Network (DCGAN)** trained to generate **anime-style faces**. The model learns patterns from real anime images and creates unique, high-quality anime faces from scratch. |
| |
|
| | --- |
| |
|
| | ## 🖼️ Sample Generated Image |
| |  |
| | <!-- <img src="https://huggingface.co/VigneshS010/anime_face_generator/smaple.png"> --> |
| |
|
| | --- |
| |
|
| | ## 📌 Model Details |
| | - **Architecture**: Deep Convolutional GAN (DCGAN) |
| | - **Dataset**: 10,000 anime images |
| | - **Framework**: TensorFlow/Keras |
| | - **File Format**: `.h5` |
| | - **Trained On**: GPU (Recommended for inference) |
| | - **Image Size**: 64 x 64 |
| |
|
| | --- |
| |
|
| | ## 🚀 How to Use the Model |
| | You can download and load the model in Python using TensorFlow/Keras: |
| |
|
| | ```python |
| | from tensorflow.keras.models import load_model |
| | |
| | # Load the model |
| | model = load_model("anime-face-generator.h5") |
| | |
| | # Generate new anime faces (example code) |
| | import numpy as np |
| | random_latent_vector = np.random.normal(size=(1, 100)) # Adjust size based on your latent space |
| | generated_image = model.predict(random_latent_vector) |
| | |
