Learning to Hallucinate Face Images via Component Generation and Enhancement
arXiv:1708.00223
Abstract
We propose a two-stage method for face hallucination. First, we generate facial components of the input image using CNNs. These components represent the basic facial structures. Second, we synthesize fine-grained facial structures from high resolution training images. The details of these structures are transferred into facial components for enhancement. Therefore, we generate facial components to approximate ground truth global appearance in the first stage and enhance them through recovering details in the second stage. The experiments demonstrate that our method performs favorably against state-of-the-art methods
arXiv commentIJCAI 2017. Project page: http://www.cs.cityu.edu.hk/~yibisong/ijcai17_sr/index.html
Cited by in corpus (8)
- SiGAN: Siamese Generative Adversarial Network for Identity-Preserving Face Hallucination
- VITAL: VIsual Tracking via Adversarial Learning
- FSRNet: End-to-End Learning Face Super-Resolution with Facial Priors
- ArtFlow: Unbiased Image Style Transfer via Reversible Neural Flows
- FCSR-GAN: Joint Face Completion and Super-resolution via Multi-task Learning
- Joint Face Hallucination and Deblurring via Structure Generation and Detail Enhancement
- Context-Patch Face Hallucination Based on Thresholding Locality-constrained Representation and Reproducing Learning
- Neural Alignment for Face De-pixelization