---
license: mit
---
# facegen-facenet-unet-gan-embedding

## Model Overview

![image/png](https://cdn-uploads.huggingface.co/production/uploads/666c3d6489e21df7d4a02805/hAhkaJMaC1s122PudZXv-.png)

This repository hosts a generative model trained to synthesize 128×128 face images conditioned on facial embeddings. The architecture jointly trains a **FaceNet encoder** and a **UNet-based GAN generator** to produce high-fidelity images from identity embeddings.

## Codebase, Dataset, and Model Artifacts

| Category         | Description                                          | Link                                                                 |
|------------------|------------------------------------------------------|----------------------------------------------------------------------|
| GitHub Repo      | Full training and inference codebase                | [GitHub Repository](https://github.com/Mayankpratapsingh022/Conditional_Face_Synthesis_with_Embedding_Conditioned_Generative_Model) |
| Dataset          | Cropped faces (128x128) for training                | [Hugging Face Dataset](https://huggingface.co/datasets/Mayank022/Cropped_Face_Dataset_128x128) |
| Trained Model    | Final GAN model with FaceNet encoder                | [Hugging Face Model](https://huggingface.co/Mayank022/facegen-facenet-unet-gan-embedding) |
| Training Notebook| End-to-end model training pipeline in Colab         | [Colab Notebook](https://colab.research.google.com/drive/16vafB_pVNk_QJpquXwxMJXNme3BCGFqS?usp=sharing) |
| Inference Notebook| Generate images from embeddings                    | [Colab Notebook](https://colab.research.google.com/drive/1Y1s7fmyVfT2jnEL9l23jmkhISNYastds?usp=sharing) |

---

## Experimental Summary

Find detailed metrics, loss trends, and inference samples in the full Weights & Biases report:

> [Weights & Biases Report – Training, Metrics, and Sample Outputs](https://api.wandb.ai/links/mayankpratapsingh0022-other/x8zkffzn)


## Use Case
The model can be used for:
- Face reconstruction from embeddings
- Conditional face generation
- Evaluating zero-shot generalization for unseen identities

## Architecture
- **Encoder**: Pretrained FaceNet (jointly fine-tuned)
- **Generator**: Modified UNet with residual upsampling blocks and conditional embedding injection at multiple resolutions
- **Embedding Injection**: Done at 8×8 resolution with skip connections

## Training Details
- **Training Time**: ~6 hours on A100
- **Dataset**: 12,000 human face images
- **Losses**: MSE + Perceptual + Adversarial
- **Optimizers**: AdamW with learning rate scheduling

## Checkpoints
- `epoch_100/`
- `epoch_200/`
- `epoch_300/` (final)

## Intended Use
This model is intended for research purposes, particularly for understanding conditional generation and face embedding interpretability.

## Limitations
- Trained on a limited dataset
- May not generalize well to non-human faces or distorted embeddings