Hallucinating faces

By Simon Baker, Takeo Kanade. 2000

668 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Media Technology, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

In Generative Models

Citations in the corpus

Survey of Hallucination in Natural Language Generation (2022, background): “1The term “hallucination” first appeared in Computer Vision (CV) in Baker and Kanade [5] and carried more positive meanings, such as superresolution [5, 112], image inpainting [48], and image synthesizing [226].”
Detecting hallucinations in large language models using semantic entropy (2024, neutral): “The term hallucination in the context of machine learning originally comes from filling in ungrounded details, either as a deliberate strategy 20 or as a reliability problem 4 .”
Super-resolution from a single image (2009, methodology): “Our approach attempts to recover at each pixel its best possible resolution increase based on its patch redundancy within and across scales.”
Face Hallucination: Theory and Practice (2007, background): “International Journal of Computer Vision 75(1), 115–134, 2007 c © 2007 Springer Science + Business Media, LLC.”
Anchored Neighborhood Regression for Fast Example-Based Super-Resolution (2013, methodology): “Most of these dictionarybased methods build on the work of Freeman et al. [8] and Baker and Kanade [2].”
FSRNet: End-to-End Learning Face Super-Resolution with Facial Priors (2017, background): “Baker and Kanade [1] proposed to learn a prior on the spatial distribution of the image gradient for frontal face images.”
Ultra-Resolving Face Images by Discriminative Generative Networks (2016, methodology): “In one of the earlier works, [1] builds the relationship between HR and LR patches using Bayesian formulation such that high-frequency details can be transferred from the dataset for face hallucination.”
Image super-resolution: The techniques, applications, and future (2016, background): “role in some specific domains such as face hallucination [173,174],”
Limits on super-resolution and how to break them (2000, background): “We can compute their Gaussian G0(Ti); : : : ; GN (Ti) and LaplacianL0(Ti); : : : ; LN (Ti) pyramids, the horizontal H0(Ti); : : : ;HN (Ti) and verticalV0(Ti); : : : ; VN (Ti) first derivatives of the Gaussian pyramids, and the horizontal H(2) 0(Ti); : : : ;H 2 N(Ti) and verticalV 2 0 (Ti); : : : ; V”
Hallucinating face by position-patch (2010, methodology): “Given the low-resolution image X L and the original high-resolution image X H that is q 2 times larger than X L , the degradation process can be expressed as [1,7] where q is a positive integer and n ð i ; j Þ represents random noise.”
GAN Prior Embedded Network for Blind Face Restoration in the Wild (2021, background): “[50, 3, 2, 36] usually solve an inverse problem based on the degradation model and handcrafted priors, which demonstrate limited performance in practice.”
Image Super-Resolution Via Sparse Representation (2010, methodology): “gradient pyramid-based prediction introduced in [13] does not directly model the face prior, and the pixels are predicted individually, causing discontinuities and artifacts.”
Hallucinating face by eigentransformation (2005, methodology): “Baker and Kanade [1], [11], [16] developed a hallucination method based on the property of face images.”
A Comprehensive Survey to Face Hallucination (2013, methodology): “2.1 Gradient-based Prior for Data Modeling Baker and Kanade (2000a) proposed the first FH algorithm.”
A two-step approach to hallucinating faces: global parametric model and local nonparametric model (2001, methodology): “To broaden the application to face hallucination, the homogeneous MRF assumption (2) has to be abandoned, leading to the work by Baker and Kanade [1].”
Global-Local Face Upsampling Network (2016, background): “In particular, upsampling a low-resolution face image to a high-resolution one has been an important problem, called by its own name of face hallucination [2].”
Coupled Kernel Embedding for Low-Resolution Face Image Recognition (2012, background): “Baker and Kanade [15] propose “face hallucination” to infer the HR face image from an input LR face image based on face priors.”
Super-resolution: a comprehensive survey (2014, methodology): “Feature Pyramids The notable work in this group was developed by Baker and Kanade [57] 1999, [71], [82], [99], and [100] for face hallucination.”