ArticlePDF Available

"KKWETC" Indian Face Database

Authors:
Manisha Satone at K. K. Wagh Institute of Engineering Education and Research
Manisha Satone
Download full-text PDF
Read full-text
Download citation

Abstract and Figures

To test face recognition algorithm developed by researchers, it is needed to have proper database. This paper describes an Indian face database 'KKWETC' of visual and thermal static images of human faces. Images were taken in uncontrolled indoor environment. Database contains 816 static visible images of 68 subjects and 150 thermal images of 50 subjects. A baseline Principal Component Analysis (PCA) face recognition algorithm was tested on both databases. Researchers can use these databases to test algorithm and compare results. Database is available to research community through the procedure described at
Figures - uploaded by Manisha Satone
Author content
All figure content in this area was uploaded by Manisha Satone
Content may be subject to copyright.
Content uploaded by Manisha Satone
Author content
All content in this area was uploaded by Manisha Satone on Dec 30, 2017
Content may be subject to copyright.
International Journal of Engineering Trends and Technology (IJETT) – Volume 54 Number 1 December 2017
ISSN: 2231-5381 http://www.ijettjournal.org Page 9
„KKWETC‟ Indian Face Database
Manisha Satone
E&TC Department, KKWIEER, Nashik, India
Abstract:
To test face recognition algorithm developed by
researchers, it is needed to have proper database.
This paper describes an Indian face database
‘KKWETC’ of visual and thermal static images of
human faces. Images were taken in uncontrolled
indoor environment. Database contains 816 static
visible images of 68 subjects and 150 thermal images
of 50 subjects. A baseline Principal Component
Analysis (PCA) face recognition algorithm was tested
on both databases. Researchers can use these
databases to test algorithm and compare results.
Database is available to research community through
the procedure described at
http://engg.kkwagh.edu.in/media/post_image/databas
e_info_website.pdf.
Keywords: Face recognition, database, thermal
images.
I. INTRODUCTION
Face recognition presents a challenging
problem in the field of image analysis and computer
vision. It is a computer application for automatically
identifying or verifying a person from an image. One
of the ways to do this is by comparing selected facial
features from the image and a facial database. It is
typically used in security systems and can be
compared to other biometrics such as
fingerprint or iris recognition systems.
The appearance of a face is largely affected by a
number of factors including identity, pose,
illumination, facial expression, age, occlusion, and
facial hair [1][2]. The development of algorithms
invariant to variations requires databases of sufficient
size that include carefully controlled variations of
these factors. While there are many databases in use
currently, the choice of an appropriate database to be
used should be made based on the task given.
Furthermore, common databases are necessary to
comparatively evaluate the algorithms. The
availability of public face databases is important for
the advancement of the field. It is therefore necessary
to create Indian face database with variations in
identity, face pose, illumination and occlusion, which
will be useful for researchers in the field of face
recognition.
Recent research has demonstrated distinct
advantages of using thermal infrared imaging for
improving face recognition performance[3][4]. While
conventional cameras
sense reflected light, thermal infrared cameras
measure emitted radiation from objects such as faces.
The human face emits thermal radiation, which can
be sensed by imaging sensors, which are sensitive in
the thermal infrared (IR) band of the electromagnetic
(EM) spectrum. Heat pattern is produced by the
temperature variations on the surface of the face. This
heat pattern can be visualized as a 2D image. Due to
the presence of highly distinctive and permanent
physiological characteristics under the facial skin,
thermal image contain important information, which
can be used for face recognition. Thermal image is
independent of ambient lighting conditions as the
thermal IR sensors only capture the heat pattern
emitted by the object. Different objects emit different
range of Infrared energy according to their
temperature and characteristics. The thermal patterns
of faces are derived primarily from the pattern of
superficial blood vessels under the skin. The vein and
tissue structure of the face is unique for each person
and, therefore, the IR images are unique. Using
unique IR image of person recognition rate can be
improved [5].
II. ‘KKWETC’DATABASE DESCRIPTION
Database contains visual images and thermal
images. Visual images are with pose variations,
illumination variation and with occlusion of 68
persons with 12 images per person, whereas thermal
database consists of images of 50 persons with
position variations in three directions.
2.1 VISUAL IMAGE DATABASE
High quality static images are taken under
uncontrolled environment using Cannon 700D
camera.
2.2.1 Frontal Images
Facial mug shots are static color images, taken in
uncontrolled indoor illumination conditions
environment. There is one mug shot per subject and
these images are labeled as person number_F.jpg (e.g.
1_F.jpg). Images are in lossless 24-bit color JPEG
format with the original size of 5184 X 3456 pixels.
These mug shot images are what you would expect to
find in a law enforcement database or when
registering to a security system. There are in total 68
frontal facial mug shot images in the database, one
per subject. Sample frontal images are shown in fig.
1.
Fig.1.Sample Frontal Images
International Journal of Engineering Trends and Technology (IJETT) – Volume 54 Number 1 December 2017
ISSN: 2231-5381 http://www.ijettjournal.org Page 10
2.1.2 Different Pose Images
This set of images was taken under the same
conditions as frontal facial mug shots. Subjects‟ poses
are −90, -45, -30, +30, +45, +90 degrees ( + right
profile, -left profile). There are 6 images per subject
in this set, which gives 408 images in total. The
images are labeled as person number_P_image
number.jpg (e.g. 001_P_01.jpg to 001_P_06.jpg for
first person). Images with different pose angles are
shown in fig. 2.
Fig.2.Sample Pose Images
2.1.3. Illumination Variation Images
Images are taken by adjusting illumination
conditions for frontal image and pose variation
images. Two images are taken for two different
illumination conditions for each pose. There are 7
images per subject for each illumination condition,
which gives 952 images in total. The images are
labeled as person number_I_image number.jpg (e.g.
001_I_01.jpg to 001_I_14.jpg for first person).
Sample illumination variation images are shown in
fig.3.
Fig.3.Sample Illumination Variation Images
2.1.4 Occluded Images
One frontal image of each person is taken by
wearing spectacle and two frontal images are taken
by wearing scarf in different ways. Total 204
occluded images are available with 3 images per
person. Two images are taken for two different
illumination conditions for each pose. The images are
labeled as person number_O_image number.jpg (e.g.
001_O_01.jpg to 001_O_03.jpg for first person).
Sample occluded images are shown in fig.4.
Fig.4.Sample Occluded Images
2.2 THERMAL IMAGES
Thermal images are captured using „Flir C2‟
camera. There is one mug shot thermal image per
subject and two images with +90 and -90 degree
positions. Thermal images were labeled as person
number_T_image number.jpg (e.g. 001_T_01.jpg).
Total thermal images are 150 with 3 images per
person. Images are in lossless 24-bit color JPEG
format with the l size of 320 X 240 pixels. Sample
thermal images are shown in fig.5.
Fig.5.Sample Thermal Images
III. RESULTS
PCA [6][7] is used to find recognition rate of
visual and thermal image databases. For testing visual
database results front image of each person is taken
for training. So total 68 images were used for training
and 816 images were used for testing. We got
recognition rate of 73 %. For thermal images, we
used 50 frontal images for training and 150 images
for testing. The recognition rate was 90 %. To
improve these recognition rate more advance
algorithms can be used.
CONCLUSION:
An attempt is made to construct face image
database for visual and thermal images. A database is
to facilitate research in human face recognition. The
key features of the database are color images, which
can be used for color processing or can be converted
to gray scale for gray scale image processing, consist
of visual as well as thermal images, database consist
of images with position variation, illumination
variation and images with occlusion. Hence, the
database developed in the present work will certainly
help researchers to develop various recognition
schemes for human face recognition.
ACKNOWLEDGEMENT
The author is grateful to sponsoring authority “Board
of college & University Development”, Savitribai
Phule, Pune university of the project and also to K K
Wagh Institute of Engineering Education and
Research(KKWIEER), Nashik where this work is
carried out. Author is also thankful to staff and
students of E&TC and Electronics department of
KKWIEER who directly or indirectly helped to
complete this project.
REFERENCES
[1] R. Chellappa, C. L. Wilson and S. Sirohey, “Human and
machine recognition of faces: a survey”, Proceedings of the
IEEE, Vol.83, No.5, 705-740, 1995.
[2] Perronnin, F., Dugelay, J.-L., 2003. “An introduction to
biometrics and face recognition”. In: Proc. IMAGE‟2003:
Learning, Understanding, Information Retrieval, Medical,
Cagliari, Italy, June.
International Journal of Engineering Trends and Technology (IJETT) – Volume 54 Number 1 December 2017
ISSN: 2231-5381 http://www.ijettjournal.org Page 11
[3] Ana M. Guzman; Mohammed Goryawala; Jin Wang;
Armando Barreto; Jean Andrian; Naphtali Rishe; Malek
Adjouadi,” Thermal Imaging as a Biometrics Approach to
Facial Signature Authentication”, IEEE Journal of
Biomedical and Health Informatics, Year: 2013, Volume: 17,
Issue: 1, Pages: 214 – 222.
[4] Gabriel Hermosilla Vigneau; José Luis Verdugo; Gonzalo
Farias Castro; Francisco Pizarro; Esteban Vera, “ Thermal
Face Recognition Under Temporal Variation Conditions”,
IEEE Access, Year: 2017, Volume: 5, Pages: 9663 – 9672.
[5] Jeffrey R. Paone; Patrick J. Flynn; P. Jonathon Philips; Kevin
W. Bowyer; Richard W. Vorder Bruegge; Patrick J. Grother;
George W. Quinn; Matthew T. Pruitt; Jason M. Grant ,”
Double Trouble: Differentiating Identical Twins by Face
Recognition”, IEEE Transactions on Information Forensics
and Security, Year: 2014, Volume: 9, Issue: 2, Pages: 285 –
295.
[6] M. Turk and A. Pentland, “Eigenfaces for recognition”, J.
Cognitive Neuroscience, Vol.3, 71-86,1991.
[7] A. Pentland, B. Moghaddam and T. Starner, “View-based and
modular eigenspaces for face recognition”, Proc. IEEE Conf.
Computer vision and Pattern Recognition, Seattle, June, 84-
91,1994.
... For visual stimuli, two previously normed and controlled face databases were used. South Asian faces were taken from the KKWETC face database (Satone, 2017), and white faces were taken from the Chicago face database (Ma et al., 2015). From each database, three female faces that were shown to display no emotional valence were picked. ...
Perception in context: How racialized identities impact speech perception
Article
Full-text available
  • Apr 2022
No PDF available ABSTRACT When available, listeners use visual cues in speech perception [Zheng and Samuel (2017)]. However, it is not clear whether racialized identities impact listeners’ judgments, and if so, to what extent everyday experiences contribute to this. American, British, and Indian English varieties were paired with white and South Asian faces to test whether listeners’ intelligibility and accentedness judgments vary as a function of the faces that they saw and the varieties that they were listening to. A prior norming study was used to assess that sentences in all varieties had similar intelligibility. Listeners in a low-diverse environment (i.e., Gainesville, USA) versus a high-diverse environment (i.e., Montreal, Canada) were recruited. Racial diversity in listeners’ social network and their language diversity [i.e., language entropy, Gullifer and Titone (2020)] were measured. Results showed that listeners’ ability to transcribe sentences (i.e., intelligibility) decreased and their accentedness judgments increased for all English varieties when speech was paired with South Asian faces. Furthermore, these effects were modulated by participants’ social network diversity and their geographic context [Kutlu et al. (2021); (2022)]. We discuss the holistic impacts of the racialization of different language varieties and the role of multilingual and diverse context effects on speech perception.
View
... For visual stimuli, two previously normed and controlled face databases were used. South Asian faces were taken from the KKWETC face database (Satone, 2017), and white faces were taken from the Chicago face database (Ma et al., 2015). From each database, three female faces that were shown to display no emotional valence were picked. ...
Does race impact speech perception? An account of accented speech in two different multilingual locales
Article
Full-text available
  • Dec 2022
Upon hearing someone’s speech, a listener can access information such as the speaker’s age, gender identity, socioeconomic status, and their linguistic background. However, an open question is whether living in different locales modulates how listeners use these factors to assess speakers’ speech. Here, an audio-visual test was used to measure whether listeners’ accentedness judgments and intelligibility (i.e., speech perception) can be modulated depending on racial information in faces that they see. American, British, and Indian English were used as three different English varieties of speech. These speech samples were presented with either a white female face or a South Asian female face. Two experiments were completed in two locales: Gainesville, Florida (USA) and Montreal, Quebec (Canada). Overall, Montreal listeners were more accurate in their transcription of sentences (i.e., intelligibility) compared to Gainesville listeners. Moreover, Gainesville listeners’ ability to transcribe the same spoken sentences decreased for all varieties when listening to speech paired with South Asian faces. However, seeing a white or a South Asian face did not impact speech intelligibility for the same spoken sentences for Montreal listeners. Finally, listeners’ accentedness judgments increased for American English and Indian English when the visual information changed from a white face to a South Asian face in Gainesville, but not in Montreal. These findings suggest that visual cues for race impact speech perception to a greater degree in locales with greater ecological diversity.
View
... Visual stimuli were collected from two previously normed and controlled face databases. For South Asian faces, pictures of two female faces were randomly selected from the KKWETC face database (Satone 2017). For White faces, two female pictures were randomly selected from the Chicago face database (Ma, Correll, and Wittenbrink 2015). ...
Now You See Me, Now You Mishear Me: Raciolinguistic accounts of speech perception in different English varieties
Preprint
  • Jan 2022
Listeners can access information about a speaker such as age, gender identity, socioeconomic status, and their linguistic background upon hearing their speech. However, it is still not clear if listeners use these factors to assess speakers’ speech. Here, an audio-visual (matched-guise) test is used to measure whether listeners’ accentedness judgments can be modulated depending on the type of face that they see. American and Indian English were used as different English varieties and presented with either a White female face or a South Asian female face. Results show that listeners’ accentedness judgments increased for Indian English compared to American English. Importantly, the increase in accentedness judgments was also observed when both American English and Indian English were presented with a South Asian face compared to a White face. These findings suggest that linguistic evaluations are modulated by non-linguistic factors and that speech perception is socially gated.
View
... For visual stimuli, we used two previously normed and controlled face databases. South Asian faces were taken from the KKWETC face database (Satone 2017). White faces were taken from the Chicago face database (Ma et al. 2015). ...
The Impact of Race on Speech Perception and Accentedness Judgements in Racially Diverse and Non-diverse Groups
Article
Full-text available
  • Nov 2021
Standard varieties are often perceived as morally superior compared with nonstandard varieties (Hill 2008). Consequently, these differences lead to ideologies that racialize nonstandard varieties (Rosa 2016), and increase the negative stereotypes towards nonstandard varieties (Giles and Watson 2013). One outlet of such stereotypes can be observed with speech intelligibility and accentedness judgements. This study examines whether seeing a White or a South Asian face impacts listeners’ perception of American, British, and Indian English and to what extent listeners’ social network diversity plays a role in predicting their perception of speech. Results indicated that intelligibility scores decreased and accentedness judgements increased for all varieties when speech was paired with South Asian faces. However, listeners with less racially diverse social networks had the highest accentedness judgements. Understanding how to account for the emergence and behavioral implications of different English varieties is a pressing question, and these results shed light on how English varieties are perceived. The implications will be discussed in light of language teaching, linguistic practices, and language research.
View
... For visual stimuli, two previously normed and controlled face databases were used. South Asian faces were taken from the KKWETC face database (Satone, 2017), and white faces were taken from the Chicago face database (Ma et al., 2015). From each database, three female faces that were shown to display no emotional valence were picked. ...
Diverse environments and their impact on accentedness judgments
Article
Full-text available
  • Oct 2020
No PDF available ABSTRACT Research shows that listeners' perceived accentedness can be mediated by visual input (Babel & Russell, 2015; McGowan, 2015; Zheng & Samuel, 2017), and can change depending on their exposure to varied speech (Baese-Berk et al., 2013). Here, we tested the impact that visual input and linguistic diversity has on listeners' perceived accentedness judgments. Two experiments were conducted: one in Gainesville (USA) and one in Montreal (Canada). While these two locations were selected for their bilingual populations, Montreal has a more diverse linguistic landscape (Gullifer & Titone, 2019). Participants completed an accentedness judgment task where they were shown either a White or a South-Asian face while listening to sentences in American, British, and Indian English. They also completed a language background questionnaire, a social network questionnaire, and executive control tasks. In an ongoing study, results show that for Gainesville, participants' perceived accentedness of all three varieties increased when the visual input changed from a White face to a South-Asian face (F(2, 66) = 33.3, p < 0.001). The same effect was not observed for listeners in Montreal (F(2,23) = 0.664, p = 0.524). These preliminary findings suggest that exposure to both different accents and racial/ethnic categories on a regular basis could impact perceived accentedness judgements.
View
... For South Asian faces, pictures of two female faces were randomly selected from the KKWETC face database (Satone, 2017). For White faces, two female pictures were randomly selected from the Chicago face database (Ma, Correll, & Wittenbrink, 2015). ...
Now You See Me, Now You Mishear Me: Raciolinguistic accounts of speech perception in different English varieties
Article
Full-text available
  • Oct 2020
Listeners can access information about a speaker such as age, gender identity, socioeconomic status, and their linguistic background upon hearing their speech. However, it is still not clear if listeners use these factors to assess speakers’ speech. Here, an audio-visual (matched-guise) test is used to measure whether listeners’ accentedness judgments can be modulated depending on the type of face that they see. American and Indian English were used as different English varieties and presented with either a White female face or a South Asian female face. Results show that listeners’ accentedness judgments increased for Indian English compared to American English. Importantly, the increase in accentedness judgments was also observed when both American English and Indian English were presented with a South Asian face compared to a White face. These findings suggest that linguistic evaluations are modulated by non-linguistic factors and that speech perception is socially gated.
View
... Asian faces were taken from the KKWETC face database (Satone 2017). White faces were taken from the Chicago face database (Ma et al. 2015 and each participant only saw one list (n=120 sentences). ...
The Impact of Race on Speech Perception and Accentedness Judgments in Racially Diverse and Non-Diverse Groups
Preprint
Full-text available
  • Aug 2020
Upon hearing someone’s speech, a listener can make inferences regarding the speakers’ age, gender identity, socioeconomic status, and their linguistic background. However, it is not clear to what extent listeners use these factors to decode the speech signal itself. Here, we use an audio-visual task to measure whether listeners’ accentedness and intelligibility judgments (i.e., speech perception) change as a function of the racial information that they see on a computer screen. American, British, and Indian English were used as three different English varieties and presented with either a White female face or a South Asian female face. Results showed that listeners’ ability to transcribe sentences (i.e., intelligibility) decreased and their accentedness judgments increased for all varieties when speech was paired with South Asian faces. However, this increase was modulated by participants’ social network diversity. In short, the racial diversity in people’s social network impacts accentedness judgments but not speech perception.
View
View
Thermal Face Recognition Under Temporal Variation Conditions
Article
Full-text available
  • May 2017
In this paper we analyze the problems produced by temporal variations of infrared face images when used in face recognition systems. The temporal variations present in thermal face images are mainly due to different environmental conditions, physiological changes of the subjects and differences of the infrared detectors’ responsivity at the time of the capture, which affect the performance of infrared face recognition systems. To perform this study, we created two thermal face databases that include capture sessions with real and variable conditions. We also propose two criteria to quantify the temporal variations between datasets. The thermal face recognition systems have been developed using the following five methods: Local Binary Pattern (LBP), Weber Linear Descriptor (WLD), Gabor Jet Descriptors (GJD), Scale Invariant Feature Transform (SIFT), and Speeded Up Robust Features (SURF). The results indicate that the local matching based methods (WLD and LBP) are mostly immune to temporal variations, which is noticeable when the face images have been acquired with a time lapse, while the rest of the methods are clearly affected and are not suitable for practical infrared face recognition.
View
Double Trouble: Differentiating Identical Twins by Face Recognition
Article
  • Feb 2014
Facial recognition algorithms should be able to operate even when similar-looking individuals are encountered, or even in the extreme case of identical twins. An experimental data set comprised of 17486 images from 126 pairs of identical twins (252 subjects) collected on the same day and 6864 images from 120 pairs of identical twins (240 subjects) with images taken a year later was used to measure the performance on seven different face recognition algorithms. Performance is reported for variations in illumination, expression, gender, and age for both the same day and cross-year image sets. Regardless of the conditions of image acquisition, distinguishing identical twins are significantly harder than distinguishing subjects who are not identical twins for all algorithms.
View
Eigenfaces for Recognition
Article
  • Jan 1991
We have developed a near-real-time computer system that can locate and track a subject's head, and then recognize the person by comparing characteristics of the face to those of known individuals. The computational approach taken in this system is motivated by both physiology and information theory, as well as by the practical requirements of near-real-time performance and accuracy. Our approach treats the face recognition problem as an intrinsically two-dimensional (2-D) recognition problem rather than requiring recovery of three-dimensional geometry, taking advantage of the fact that faces are normally upright and thus may be described by a small set of 2-D characteristic views. The system functions by projecting face images onto a feature space that spans the significant variations among known face images. The significant features are known as "eigenfaces," because they are the eigenvectors (principal components) of the set of faces; they do not necessarily correspond to features such as eyes, ears, and noses. The projection operation characterizes an individual face by a weighted sum of the eigenface features, and so to recognize a particular face it is necessary only to compare these weights to those of known individuals. Some particular advantages of our approach are that it provides for the ability to learn and later recognize new faces in an unsupervised manner, and that it is easy to implement using a neural network architecture.
View
Thermal Imaging as a Biometrics Approach to Facial Signature Authentication
Article
  • Jul 2012
  • IEEE T INF TECHNOL B
A new thermal imaging framework with unique feature extraction and similarity measurements for face recognition is presented. The research premise is to design specialized algorithms that would extract vasculature information, create a thermal facial signature and identify the individual. The proposed algorithm is fully integrated and consolidates the critical steps of feature extraction through the use of morphological operators, registration using the Linear Image Registration Tool (FLIRT) and matching through unique similarity measures designed for this task. The novel approach at developing a thermal signature template using four images taken at various instants of time ensured that unforeseen changes in the vasculature over time did not affect the biometric matching process as the authentication process relied only on consistent thermal features. Thirteen subjects were used for testing the developed technique on an in-house thermal imaging system. The matching using the similarity measures showed an average accuracy of 88.46% for skeletonized signatures and 90.39% for anisotropically diffused signatures. The highly accurate results obtained in the matching process clearly demonstrate the ability of the thermal infrared system to extend in application to other thermal imaging based systems. Empirical results applying this approach to an existing database of thermal images proves this assertion.
View
Human and Machine Recognition of Faces: A Survey
Article
  • Jun 1995
The goal of this paper is to present a critical survey of existing literature on human and machine recognition of faces. Machine recognition of faces has several applications, ranging from static matching of controlled photographs as in mug shots matching and credit card verification to surveillance video images. Such applications have different constraints in terms of complexity of processing requirements and thus present a wide range of different technical challenges. Over the last 20 years researchers in psychophysics, neural sciences and engineering, image processing analysis and computer vision have investigated a number of issues related to face recognition by humans and machines. Ongoing research activities have been given a renewed emphasis over the last five years. Existing techniques and systems have been tested on different sets of images of varying complexities. But very little synergism exists between studies in psychophysics and the engineering literature. Most importantly, there exists no evaluation or benchmarking studies using large databases with the image quality that arises in commercial and law enforcement applications In this paper, we first present different applications of face recognition in commercial and law enforcement sectors. This is followed by a brief overview of the literature on face recognition in the psychophysics community. We then present a detailed overview of move than 20 years of research done in the engineering community. Techniques for segmentation/location of the face, feature extraction and recognition are reviewed. Global transform and feature based methods using statistical, structural and neural classifiers are summarized
View
View-Based and Modular Eigenspaces for Face Recognition
Article
  • Oct 1994
  • IEEE Comput Soc Conf Comput Vis Pattern Recogn
In this work we describe experiments with eigenfaces for recognition and interactive search in a large-scale face database. Accurate visual recognition is demonstrated using a database of O(10 3 ) faces. The problem of recognition under general viewing orientation is also examined. A view-based multiple-observer eigenspace technique is proposed for use in face recognition under variable pose. In addition, a modular eigenspace description technique is used which incorporates salient features such as the eyes, nose and mouth, in an eigenfeature layer. This modular representation yields higher recognition rates as well as a more robust framework for face recognition. An automatic feature extraction technique using feature eigentemplates is also demonstrated. 1 Introduction In recent years considerable progress has been made on the problems of face detection and recognition, especially in the processing of "mug shots," i.e., head-on face pictures with controlled illumination and scale...
View
An introduction to biometrics and face recognition
  • Jun 2003
  • 2231-5381
  • F Perronnin
  • J.-L Dugelay
Perronnin, F., Dugelay, J.-L., 2003. "An introduction to biometrics and face recognition". In: Proc. IMAGE"2003: Learning, Understanding, Information Retrieval, Medical, Cagliari, Italy, June. ISSN: 2231-5381 http://www.ijettjournal.org Page 11
José Luis Verdugo; Gonzalo Farias Castro; Francisco Pizarro Thermal Face Recognition Under Temporal Variation Conditions
  • 2017-9663
  • Gabriel Hermosilla
  • Vigneau
Gabriel Hermosilla Vigneau; José Luis Verdugo; Gonzalo Farias Castro; Francisco Pizarro; Esteban Vera, " Thermal Face Recognition Under Temporal Variation Conditions", IEEE Access, Year: 2017, Volume: 5, Pages: 9663-9672.