Face Hallucination: Theory and Practice

@article{Liu2006FaceHT,
  title={Face Hallucination: Theory and Practice},
  author={Ce Liu and Harry Shum and William T. Freeman},
  journal={International Journal of Computer Vision},
  year={2006},
  volume={75},
  pages={115-134}
}
In this paper, we study face hallucination, or synthesizing a high-resolution face image from an input low-resolution image, with the help of a large collection of other high-resolution face images. [] Key Method At the first step, we derive a global linear model to learn the relationship between the high-resolution face images and their smoothed and down-sampled lower resolution ones.
A two-step face hallucination approach for video surveillance applications
TLDR
A novel face hallucination algorithm to synthesize a high-resolution face image from several low-resolution input face images and uses a sparse representation and the ANN method to enhance both global face shape and local high frequency information while greatly improving the processing speed.
Hallucinating face by position-patch
Face Hallucination Using Cascaded Super-Resolution and Identity Priors
TLDR
The proposed C-SRIP model (Cascaded Super Resolution with Identity Priors) is able to upscale (tiny) low- resolution images captured in unconstrained conditions and produce visually convincing results for diverse low-resolution inputs.
Fast face hallucination with sparse representation for video surveillance
TLDR
This paper proposes a novel face hallucination algorithm to synthesize a high-resolution face image from several low-resolution input face images, and improves the example-based super-resolution method for local high frequency information enhancement.
Structured Face Hallucination
TLDR
Experimental results demonstrate that the proposed algorithm generates hallucinated face images with favorable quality and adaptability.
Face hallucination through ensemble learning
TLDR
The experimental results show that the proposed framework is capable of synthesizing high-resolution images from low-resolution input images with a wide variety of facial poses, geometry misalignments and facial expressions even when such images are not included within the original training dataset.
Face hallucination based on morphological component analysis
Face Hallucination via Similarity Constraints
TLDR
A new face hallucination method based on similarity constraints to produce a high-resolution face image from an input low-resolution (LR) face image by incorporating four constraint functions at patch level.
Face hallucination based on two-dimensional joint learning
...
...

References

SHOWING 1-10 OF 53 REFERENCES
A two-step approach to hallucinating faces: global parametric model and local nonparametric model
  • Ce LiuH. ShumChangshui Zhang
  • Computer Science
    Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001
  • 2001
TLDR
A two-step statistical modeling approach that integrates both a global parametric model and a local nonparametric model is developed, which can generate photorealistic face images.
Hallucinating face by eigentransformation
TLDR
Experiments show that the hallucinated face images are not only very helpful for recognition by humans, but also make the automatic recognition procedure easier, since they emphasize the face difference by adding more high-frequency details.
Face hallucination with pose variation
  • Yang LiX. Lin
  • Computer Science
    Sixth IEEE International Conference on Automatic Face and Gesture Recognition, 2004. Proceedings.
  • 2004
TLDR
A texture model is derived consisting of a set of linear mappings between the Gabor wavelet features of the facial images of every two possible poses that can be used to synthesize a high-resolution facial image from a low-resolution input.
Multi-modal tensor face for simultaneous super-resolution and recognition
  • K. JiaS. Gong
  • Computer Science
    Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1
  • 2005
TLDR
This paper presents a Bayesian framework to perform multimodal (such as variations in viewpoint and illumination) face image super-resolution for recognition in tensor space, and integrates the tasks of super- resolution and recognition by directly computing a maximum likelihood identity parameter vector in high-resolution Tensor space for recognition.
Super-resolution from multiple views using learnt image models
  • David P. CapelAndrew Zisserman
  • Mathematics
    Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001
  • 2001
The objective of the work presented is the super-resolution restoration of a set of images, and we investigate the use of learnt image models within a generative Bayesian framework. It is
Resolution-Aware Fitting of Active Appearance Models to Low Resolution Images
TLDR
Experimental results show that RAF considerably improves the estimation accuracy of both shape and appearance parameters when fitting to low resolution data, and is compared against a state-of-the-art tracker.
Image analogies
TLDR
This paper describes a new framework for processing images by example, called “image analogies,” based on a simple multi-scale autoregression, inspired primarily by recent results in texture synthesis.
Robust multipose face detection in images
TLDR
This work proposes a novel three-step face detection approach that greatly improves detection accuracy with small computation cost and is shown to be more effective and capable of handling more pose variations than conventional approaches.
Lucas-Kanade 20 Years On: A Unifying Framework
TLDR
An overview of image alignment is presented, describing most of the algorithms and their extensions in a consistent framework and concentrating on the inverse compositional algorithm, an efficient algorithm that was recently proposed.
Rotation Invariant Neural Network-Based Face Detection
  • H. RowleyS. BalujaT. Kanade
  • Computer Science
    Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231)
  • 1998
TLDR
A neural network is used to analyze each window of the input before it is processed by a “detector” network, which decides whether a face is present, which is significantly faster than exhaustively trying all orientations, and will yield fewer false detections.
...
...