Cross-spectrum thermal to visible face recognition based on cascaded image synthesis

Mallat, Khawla; Damer, Naser; Boutros, Fadi; Kuijper, Arjan; Dugelay, Jean‐Luc

doi:10.1109/icb45273.2019.8987347

Cited by 32 publications

(23 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Source: [4] A common triangulation of both reference shape and detected facial landmarks makes it possible to compute a piecewise affine transformation to each triangle and apply a set of transformations to transform a face from an arbitrary position in the image into a well-defined coordinate system. This makes it possible to use fixed regions of interest (ROIs) for image analysis, even for moving faces, so additionally increasing the number of algorithms that can be applied to images with unconstrained movement.…”

Section: Figure 2: Samples Of Generated Images Acquired In Total Darkmentioning

confidence: 99%

“…The face detector mainly uses the bimodal temperature distribution of human skin and typical indoor backgrounds. Mallat et al discussed a novel solution based on cascaded refinement networks, which was able to generate high-quality color visible images, trained on a limited size database [ 4 ]. Their network is based on the use of contextual loss functions, enabling it to be inherently scale and rotation invariant.…”

Section: Reviewmentioning

confidence: 99%

“…Poor or absent illumination (shown in the right column) posed no impact on the images generated. Synthesizing images with informative facial attributes that are not in the visible spectrum was achieved [ 4 ].…”

Section: Reviewmentioning

confidence: 99%

See 2 more Smart Citations

Thermal Infrared Face Recognition

Weidlich¹

2021

Cureus

View full text Add to dashboard Cite

The technology for deep learning in the field of thermal infrared face recognition has recently become more available for use in research, therefore allowing for the many groups working on this subject to achieve many novel findings. Thermal infrared face recognition helps recognize faces that are not able to be recognized in visible light and can additionally recognize facial blood vessel structure. Previous research regarding temperature variations, mathematical formulas, wave types, and methods in thermal infrared face recognition is reviewed.

show abstract

Section: Figure 2: Samples Of Generated Images Acquired In Total Darkmentioning

confidence: 99%

Section: Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Thermal Infrared Face Recognition

Weidlich¹

2021

Cureus

View full text Add to dashboard Cite

show abstract

“…More recently, [7] introduced a method that incorporates facial attributes by pooling latent features with attribute features and synthesizes visible domain images at multiple scales to guide face synthesis effectively. Similarly, [20] considers multi-scale information for higher resolution generation with less training data using a series of cascade refinement networks.…”

Section: Related Workmentioning

confidence: 99%

“…AUC ↑ EER ↓ TAR@1% ↑ TAR@5% ↑ GANVFS [31] 73.8 32.3 --CRN + CL [20] 74.9 31.7 --Multi-AP-GAN [7] The TUFTS face dataset is challenging due to the limited number of training examples. Despite this limitation, Table III shows that our proposed method similarly improves on all reported metrics.…”

Section: Modelmentioning

confidence: 99%

A Synthesis-Based Approach for Thermal-to-Visible Face Verification

Peri¹,

Gleason²,

Castillo³

et al. 2021

Preprint

View full text Add to dashboard Cite

In recent years, visible-spectrum face verification systems have been shown to match expert forensic examiner recognition performance. However, such systems are ineffective in low-light and nighttime conditions. Thermal face imagery, which captures body heat emissions, effectively augments the visible spectrum, capturing discriminative facial features in scenes with limited illumination. Due to the increased cost and difficulty of obtaining diverse, paired thermal and visible spectrum datasets, algorithms and large-scale benchmarks for low-light recognition are limited. This paper presents an algorithm that achieves state-of-the-art performance on both the ARL-VTF and TUFTS multi-spectral face datasets. Importantly, we study the impact of face alignment, pixel-level correspondence, and identity classification with label smoothing for multi-spectral face synthesis and verification. We show that our proposed method is widely applicable, robust, and highly effective. In addition, we show that the proposed method significantly outperforms face frontalization methods on profileto-frontal verification. Finally, we present MILAB-VTF(B), a challenging multi-spectral face dataset that is composed of paired thermal and visible videos. To the best of our knowledge, with face data from 400 subjects, this dataset represents the most extensive collection of publicly available indoor and longrange outdoor thermal-visible face imagery. Lastly, we show that our end-to-end thermal-to-visible face verification system provides strong performance on the MILAB-VTF(B) dataset.

show abstract

Thermal Face Recognition Based on Multi-scale Image Synthesis

Chu

Huang

2021

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Cross-spectrum thermal to visible face recognition based on cascaded image synthesis

Cited by 32 publications

References 16 publications

Thermal Infrared Face Recognition

Thermal Infrared Face Recognition

A Synthesis-Based Approach for Thermal-to-Visible Face Verification

Thermal Face Recognition Based on Multi-scale Image Synthesis

Contact Info

Product

Resources

About