Look globally, age locally: Face aging with an attention mechanism (AcGANs)

PyTorch implementation of the AcGANs algorithm in the paper ``Look globally, age locally: Face aging with an attention mechanism.''.

1. The Architecture of AcGANs

---

2. Prerequisites

---

• Python 3.6

• PyTorch 1.3.0

• GPU

3. Dataset & Preparation

---

• Morph
• CACD

4. Training

---

Training a model by:

$ python main.py config/morph.yml

5. Results

---

• Attention Results

• 

• Results on the Morph Dataset

• 

• Comparison of AcGANs, IPCGANs, and CAAE in the Morph Dataset

  

6. Citation

---

Zhu H, Huang Z, Shan H, et al. Look globally, age locally: Face aging with an attention mechanism[J]. arXiv preprint arXiv:1910.12771, 2019.

7. License

---

AcGANs is freely available for free non-commercial use, and may be redistributed under these conditions. For commercial queries, contact Junping Zhang.