Jeremy Dawson
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jeremy Dawson has a total of 78 co-authors. Here are the top co-authors by number of shared publications:
- Nasser M. Nasrabadi
West Virginia University
21 shared publications Ali Dabouei
West Virginia University
11 shared publications- Sobhan Soleymani
Lane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WV, USA
9 shared publications - Dimitris Korakakis
West Virginia University, Lane Department of Computer Science and Electrical Engineering, Morgantown, West Virginia, United States
3 shared publications Fariborz Taherkhani
Department of Computer Science and Electrical Engineering West Virginia University
3 shared publications- Hadi Kazemi
West Virginia University
3 shared publications
Latest Publications
- Open Access
Finger Type Classification for Fingerprint Image Error Correction in Large Scale Biometric Databases
Large-scale biometric systems, essential for national security and border management, increasingly rely on multimodal databases containing millions of identities. However, operational pressures and insufficient training lead to frequent image classification and labeling errors by human operators. These critical data integrity issues include the mislabeling of rolled vs. flat fingerprints, out-of-sequence captures, and the insertion of incorrect modalities. Such errors render enrollment records unreliable, compromising subsequent identity verification processes. Since manually sorting vast image archives is unfeasible, our study proposes an automated solution. The primary objective was to deploy a Siamese Network to classify fingerprints by their precise finger type and collection methodology (flat or rolled impressions). A secondary, but central, goal was to investigate the influence of varying embedding dimensions (64, 128, 256, 512) and similarity thresholds (0.5, 0.2, 0.1) on the network’s performance metrics. Our most significant finding demonstrates a clear trade-off: a lower similarity threshold drastically increases conditional accuracy and precision (e.g., up to 98%) but simultaneously increases the proportion of images categorized as "uncertain" (up to 24%). In a practical, large-scale application, this necessitates balancing superior classification accuracy against a higher volume of images requiring costly manual inspection. This work provides a proof-of-concept tool capable of efficiently quantifying the percentage of images requiring human review across various modalities (fingerprints, face, iris). The eventual goal is a lightweight, efficient tool to establish standard preprocessing procedures for any large biometric dataset, dramatically reducing the time and cost associated with data integrity maintenance.
Open AccessFinger Type Classification for Fingerprint Image Error Correction in Large Scale Biometric Databases
Large-scale biometric systems, essential for national security and border management, increasingly rely on multimodal databases containing millions of identities. However, operational pressures and insufficient training lead to frequent image classification and labeling errors by human operators. These critical data integrity issues include the mislabeling of rolled vs. flat fingerprints, out-of-sequence captures, and the insertion of incorrect modalities. Such errors render enrollment records unreliable, compromising subsequent identity verification processes. Since manually sorting vast image archives is unfeasible, our study proposes an automated solution. The primary objective was to deploy a Siamese Network to classify fingerprints by their precise finger type and collection methodology (flat or rolled impressions). A secondary, but central, goal was to investigate the influence of varying embedding dimensions (64, 128, 256, 512) and similarity thresholds (0.5, 0.2, 0.1) on the network’s performance metrics. Our most significant finding demonstrates a clear trade-off: a lower similarity threshold drastically increases conditional accuracy and precision (e.g., up to 98%) but simultaneously increases the proportion of images categorized as "uncertain" (up to 24%). In a practical, large-scale application, this necessitates balancing superior classification accuracy against a higher volume of images requiring costly manual inspection. This work provides a proof-of-concept tool capable of efficiently quantifying the percentage of images requiring human review across various modalities (fingerprints, face, iris). The eventual goal is a lightweight, efficient tool to establish standard preprocessing procedures for any large biometric dataset, dramatically reducing the time and cost associated with data integrity maintenance.
- Open Access
WaFusion: A Wavelet-Enhanced Diffusion Framework for Face Morph Generation
Biometric face morphing poses a critical challenge to identity verification systems, undermining their security and robustness. To address this issue, we propose WaFusion, a novel framework combining wavelet decomposition and diffusion models to generate high-quality, realistic morphed face images efficiently. WaFusion leverages the structural details captured by wavelet transforms and the generative capabilities of diffusion models, producing face morphs with minimal artifacts. Experiments conducted on FERET, FRGC, FRLL, and WVU Twin datasets demonstrate WaFusion's superiority over state-of-the-art methods, producing high-resolution morphs with fewer artifacts. Our framework excels across key biometric metrics, including the Attack Presentation Classification Error Rate (APCER), Bona Fide Presentation Classification Error Rate (BPCER), and Equal Error Rate (EER). This work sets a new benchmark in biometric morph generation, offering a cutting-edge and efficient solution to enhance biometric security systems.
Open AccessWaFusion: A Wavelet-Enhanced Diffusion Framework for Face Morph Generation
Biometric face morphing poses a critical challenge to identity verification systems, undermining their security and robustness. To address this issue, we propose WaFusion, a novel framework combining wavelet decomposition and diffusion models to generate high-quality, realistic morphed face images efficiently. WaFusion leverages the structural details captured by wavelet transforms and the generative capabilities of diffusion models, producing face morphs with minimal artifacts. Experiments conducted on FERET, FRGC, FRLL, and WVU Twin datasets demonstrate WaFusion's superiority over state-of-the-art methods, producing high-resolution morphs with fewer artifacts. Our framework excels across key biometric metrics, including the Attack Presentation Classification Error Rate (APCER), Bona Fide Presentation Classification Error Rate (BPCER), and Equal Error Rate (EER). This work sets a new benchmark in biometric morph generation, offering a cutting-edge and efficient solution to enhance biometric security systems.
- Open Access
GIF: Generative Inspiration for Face Recognition at Scale
Aiming to reduce the computational cost of Softmax in massive label space of Face Recognition (FR) benchmarks, recent studies estimate the output using a subset of identities. Although promising, the association between the computation cost and the number of identities in the dataset remains linear only with a reduced ratio. A shared characteristic among available FR methods is the employment of atomic scalar labels during training. Consequently, the input to label matching is through a dot product between the feature vector of the input and the Softmax centroids. Inspired by generative modeling, we present a simple yet effective method that substitutes scalar labels with structured identity code, i.e., a sequence of integers. Specifically, we propose a tokenization scheme that transforms atomic scalar labels into structured identity codes. Then, we train an FR backbone to predict the code for each input instead of its scalar label. As a result, the associated computational cost becomes logarithmic w.r.t. number of identities. We demonstrate the benefits of the proposed method by conducting experiments. In particular, our method outperforms its competitors by 1.52%, and 0.6% at TAR@FAR$=1e-4$ on IJB-B and IJB-C, respectively, while transforming the association between computational cost and the number of identities from linear to logarithmic. See code at https://github.com/msed-Ebrahimi/GIF
Open AccessGIF: Generative Inspiration for Face Recognition at Scale
Aiming to reduce the computational cost of Softmax in massive label space of Face Recognition (FR) benchmarks, recent studies estimate the output using a subset of identities. Although promising, the association between the computation cost and the number of identities in the dataset remains linear only with a reduced ratio. A shared characteristic among available FR methods is the employment of atomic scalar labels during training. Consequently, the input to label matching is through a dot product between the feature vector of the input and the Softmax centroids. Inspired by generative modeling, we present a simple yet effective method that substitutes scalar labels with structured identity code, i.e., a sequence of integers. Specifically, we propose a tokenization scheme that transforms atomic scalar labels into structured identity codes. Then, we train an FR backbone to predict the code for each input instead of its scalar label. As a result, the associated computational cost becomes logarithmic w.r.t. number of identities. We demonstrate the benefits of the proposed method by conducting experiments. In particular, our method outperforms its competitors by 1.52%, and 0.6% at TAR@FAR$=1e-4$ on IJB-B and IJB-C, respectively, while transforming the association between computational cost and the number of identities from linear to logarithmic. See code at https://github.com/msed-Ebrahimi/GIF
- 23 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biometrics, Behavior, and Identity Science
- Vol. 7 (2) , 195-209
Learning Multi-Scale Knowledge-Guided Features for Text-Guided Face Recognition
Text-guided face recognition (TGFR) aims to improve the performance of state-of-the-art face recognition (FR) algorithms by incorporating auxiliary information, such as distinct facial marks and attributes, provided as natural language descriptions. Current TGFR algorithms have been proven to be highly effective in addressing performance drops in state-of-the-art FR models, particularly in scenarios involving sensor noise, low resolution, and turbulence effects. Although existing methods explore various algorithms using different cross-modal alignment and fusion techniques, they encounter practical limitations in real-world applications. For example, during inference, textual descriptions associated with face images may be missing, lacking crucial details, or incorrect. Furthermore, the presence of inherent modality heterogeneity poses a significant challenge in achieving effective cross-modal alignment. To address these challenges, we introduce CaptionFace, a TGFR framework that integrates GPTFace, a face image captioning model designed to generate context-rich natural language descriptions from low-resolution facial images. By leveraging GPTFace, we overcome the issue of missing textual descriptions, expanding the applicability of CaptionFace to single-modal FR datasets. Additionally, we introduce a multi-scale feature alignment (MSFA) module to ensure semantic alignment between face-caption pairs at different granularities. Furthermore, we introduce an attribute-aware loss and perform knowledge adaptation to specifically adapt textual knowledge from facial features. Extensive experiments on three face-caption datasets and various unconstrained single-modal benchmark datasets demonstrate that CaptionFace significantly outperforms state-of-the-art FR models and existing TGFR approaches.
- 23 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Biometrics, Behavior, and Identity Science
- Vol. 7 (2) , 195-209
Learning Multi-Scale Knowledge-Guided Features for Text-Guided Face Recognition
Text-guided face recognition (TGFR) aims to improve the performance of state-of-the-art face recognition (FR) algorithms by incorporating auxiliary information, such as distinct facial marks and attributes, provided as natural language descriptions. Current TGFR algorithms have been proven to be highly effective in addressing performance drops in state-of-the-art FR models, particularly in scenarios involving sensor noise, low resolution, and turbulence effects. Although existing methods explore various algorithms using different cross-modal alignment and fusion techniques, they encounter practical limitations in real-world applications. For example, during inference, textual descriptions associated with face images may be missing, lacking crucial details, or incorrect. Furthermore, the presence of inherent modality heterogeneity poses a significant challenge in achieving effective cross-modal alignment. To address these challenges, we introduce CaptionFace, a TGFR framework that integrates GPTFace, a face image captioning model designed to generate context-rich natural language descriptions from low-resolution facial images. By leveraging GPTFace, we overcome the issue of missing textual descriptions, expanding the applicability of CaptionFace to single-modal FR datasets. Additionally, we introduce a multi-scale feature alignment (MSFA) module to ensure semantic alignment between face-caption pairs at different granularities. Furthermore, we introduce an attribute-aware loss and perform knowledge adaptation to specifically adapt textual knowledge from facial features. Extensive experiments on three face-caption datasets and various unconstrained single-modal benchmark datasets demonstrate that CaptionFace significantly outperforms state-of-the-art FR models and existing TGFR approaches.
- Open Access
FDWST: Fingerphoto Deblurring using Wavelet Style Transfer
The challenge of deblurring fingerphoto images, or generating a sharp fingerphoto from a given blurry one, is a significant problem in the realm of computer vision. To address this problem, we propose a fingerphoto deblurring architecture referred to as Fingerphoto Deblurring using Wavelet Style Transfer (FDWST), which aims to utilize the information transmission of Style Transfer techniques to deblur fingerphotos. Additionally, we incorporate the Discrete Wavelet Transform (DWT) for its ability to split images into different frequency bands. By combining these two techniques, we can perform Style Transfer over a wide array of wavelet frequency bands, thereby increasing the quality and variety of sharpness information transferred from sharp to blurry images. Using this technique, our model was able to drastically increase the quality of the generated fingerphotos compared to their originals, and achieve a peak matching accuracy of 0.9907 when tasked with matching a deblurred fingerphoto to its sharp counterpart, outperforming multiple other state-of-the-art deblurring and style transfer techniques.
Open AccessFDWST: Fingerphoto Deblurring using Wavelet Style Transfer
The challenge of deblurring fingerphoto images, or generating a sharp fingerphoto from a given blurry one, is a significant problem in the realm of computer vision. To address this problem, we propose a fingerphoto deblurring architecture referred to as Fingerphoto Deblurring using Wavelet Style Transfer (FDWST), which aims to utilize the information transmission of Style Transfer techniques to deblur fingerphotos. Additionally, we incorporate the Discrete Wavelet Transform (DWT) for its ability to split images into different frequency bands. By combining these two techniques, we can perform Style Transfer over a wide array of wavelet frequency bands, thereby increasing the quality and variety of sharpness information transferred from sharp to blurry images. Using this technique, our model was able to drastically increase the quality of the generated fingerphotos compared to their originals, and achieve a peak matching accuracy of 0.9907 when tasked with matching a deblurred fingerphoto to its sharp counterpart, outperforming multiple other state-of-the-art deblurring and style transfer techniques.
- Open Access
Synthetic Latent Fingerprint Generation Using Style Transfer
Limited data availability is a challenging problem in the latent fingerprint domain. Synthetically generated fingerprints are vital for training data-hungry neural network-based algorithms. Conventional methods distort clean fingerprints to generate synthetic latent fingerprints. We propose a simple and effective approach using style transfer and image blending to synthesize realistic latent fingerprints. Our evaluation criteria and experiments demonstrate that the generated synthetic latent fingerprints preserve the identity information from the input contact-based fingerprints while possessing similar characteristics as real latent fingerprints. Additionally, we show that the generated fingerprints exhibit several qualities and styles, suggesting that the proposed method can generate multiple samples from a single fingerprint.
Open AccessSynthetic Latent Fingerprint Generation Using Style Transfer
Limited data availability is a challenging problem in the latent fingerprint domain. Synthetically generated fingerprints are vital for training data-hungry neural network-based algorithms. Conventional methods distort clean fingerprints to generate synthetic latent fingerprints. We propose a simple and effective approach using style transfer and image blending to synthesize realistic latent fingerprints. Our evaluation criteria and experiments demonstrate that the generated synthetic latent fingerprints preserve the identity information from the input contact-based fingerprints while possessing similar characteristics as real latent fingerprints. Additionally, we show that the generated fingerprints exhibit several qualities and styles, suggesting that the proposed method can generate multiple samples from a single fingerprint.
- Open Access
- 2 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 82709-82727
Fingerphoto Deblurring Using Attention-Guided Multi-Stage GAN
Using fingerphoto images acquired from mobile cameras, low-quality sensors, or crime scenes, it has become a challenge for automated identification systems to verify the identity due to various acquisition distortions. A significant type of photometric distortion that notably reduces the quality of a fingerphoto is the blurring of the image. This paper proposes a deep fingerphoto deblurring model to restore the ridge information degraded by the image blurring. As the core of our model, we utilize a conditional Generative Adversarial Network (cGAN) to learn the distribution of natural ridge patterns. We perform several modifications to enhance the quality of the reconstructed (deblurred) fingerphotos by our proposed model. First, we develop a multi-stage GAN to learn the ridge distribution in a coarse-to-fine framework. This framework enables the model to maintain the consistency of the ridge deblurring process at different resolutions. Second, we propose a guided attention module that helps the generator to focus mainly on blurred regions. Third, we incorporate a deep fingerphoto verifier as an auxiliary adaptive loss function to force the generator to preserve the ID information during the deblurring process. Finally, we evaluate the effectiveness of the proposed model through extensive experiments on multiple public fingerphoto datasets as well as real-world blurred fingerphotos. In particular, our method achieves 5.2 dB, 8.7%, and 7.6% improvement in PSNR, AUC, and EER, respectively, compared to a state-of-the-art deblurring method.
- 2 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 82709-82727
Open AccessFingerphoto Deblurring Using Attention-Guided Multi-Stage GAN
Using fingerphoto images acquired from mobile cameras, low-quality sensors, or crime scenes, it has become a challenge for automated identification systems to verify the identity due to various acquisition distortions. A significant type of photometric distortion that notably reduces the quality of a fingerphoto is the blurring of the image. This paper proposes a deep fingerphoto deblurring model to restore the ridge information degraded by the image blurring. As the core of our model, we utilize a conditional Generative Adversarial Network (cGAN) to learn the distribution of natural ridge patterns. We perform several modifications to enhance the quality of the reconstructed (deblurred) fingerphotos by our proposed model. First, we develop a multi-stage GAN to learn the ridge distribution in a coarse-to-fine framework. This framework enables the model to maintain the consistency of the ridge deblurring process at different resolutions. Second, we propose a guided attention module that helps the generator to focus mainly on blurred regions. Third, we incorporate a deep fingerphoto verifier as an auxiliary adaptive loss function to force the generator to preserve the ID information during the deblurring process. Finally, we evaluate the effectiveness of the proposed model through extensive experiments on multiple public fingerphoto datasets as well as real-world blurred fingerphotos. In particular, our method achieves 5.2 dB, 8.7%, and 7.6% improvement in PSNR, AUC, and EER, respectively, compared to a state-of-the-art deblurring method.
- Open Access
- 24 July 2023
- Published by Institution of Engineering and Technology (IET) in IET Biometrics
- Vol. 12 (4) , 194-210
On improving interoperability for cross‐domain multi‐finger fingerprint matching using coupled adversarial learning
Improving interoperability in contactless‐to‐contact fingerprint matching is a crucial factor for the mainstream adoption of contactless fingerphoto devices. However, matching contactless probe images against legacy contact‐based gallery images is very challenging due to the presence of heterogeneity between these domains. Moreover, unconstrained acquisition of fingerphotos produces perspective distortion. Therefore, direct matching of fingerprint features suffers severe performance degradation on cross‐domain interoperability. In this study, to address this issue, the authors propose a coupled adversarial learning framework to learn a fingerprint representation in a low‐dimensional subspace that is discriminative and domain‐invariant in nature. In fact, using a conditional coupled generative adversarial network, the authors project both the contactless and the contact‐based fingerprint into a latent subspace to explore the hidden relationship between them using class‐specific contrastive loss and ArcFace loss. The ArcFace loss ensures intra‐class compactness and inter‐class separability, whereas the contrastive loss minimises the distance between the subspaces for the same finger. Experiments on four challenging datasets demonstrate that our proposed model outperforms state‐of‐the methods and two top‐performing commercial‐off‐the‐shelf SDKs, that is, Verifinger v12.0 and Innovatrics. In addition, the authors also introduce a multi‐finger score fusion network that significantly boosts interoperability by effectively utilising the multi‐finger input of the same subject for both cross‐domain and cross‐sensor settings.
- 24 July 2023
- Published by Institution of Engineering and Technology (IET) in IET Biometrics
- Vol. 12 (4) , 194-210
Open AccessOn improving interoperability for cross‐domain multi‐finger fingerprint matching using coupled adversarial learning
Improving interoperability in contactless‐to‐contact fingerprint matching is a crucial factor for the mainstream adoption of contactless fingerphoto devices. However, matching contactless probe images against legacy contact‐based gallery images is very challenging due to the presence of heterogeneity between these domains. Moreover, unconstrained acquisition of fingerphotos produces perspective distortion. Therefore, direct matching of fingerprint features suffers severe performance degradation on cross‐domain interoperability. In this study, to address this issue, the authors propose a coupled adversarial learning framework to learn a fingerprint representation in a low‐dimensional subspace that is discriminative and domain‐invariant in nature. In fact, using a conditional coupled generative adversarial network, the authors project both the contactless and the contact‐based fingerprint into a latent subspace to explore the hidden relationship between them using class‐specific contrastive loss and ArcFace loss. The ArcFace loss ensures intra‐class compactness and inter‐class separability, whereas the contrastive loss minimises the distance between the subspaces for the same finger. Experiments on four challenging datasets demonstrate that our proposed model outperforms state‐of‐the methods and two top‐performing commercial‐off‐the‐shelf SDKs, that is, Verifinger v12.0 and Innovatrics. In addition, the authors also introduce a multi‐finger score fusion network that significantly boosts interoperability by effectively utilising the multi‐finger input of the same subject for both cross‐domain and cross‐sensor settings.
- Open Access
- 9 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 24281-24298
Attention Augmented Face Morph Detection
Morph detection is of paramount significance when the integrity of Automatic Face Recognition (AFR) systems is concerned. Considering the risks incurred by morphing attacks, a robust automated morph detector is required which can distinguish authentic bona fide samples from altered morphed images. We leverage the wavelet sub-band decomposition of an input image, yielding the fine-grained spatial-frequency content of the input image. To enhance detection of morphed images, our goal is to find the most discriminative information across frequency channels and spatial domain. To this end, we propose an end-to-end attention-based deep morph detector which assimilates the most discriminative wavelet sub-bands of a given image which are obtained by a group sparsity representation learning scheme. Specifically, our group sparsity-constrained Deep Neural Network (DNN) learns the most discriminative wavelet sub-bands (channels) of an input image while the attention mechanism captures the most discriminative spatial regions of input images for the downstream task of morph detection. To this end, we adopt three attention mechanisms to diversify our refined features for morph detection. As the first attention mechanism, we employ the Convolutional Block Attention Module (CBAM) which provides us with refined feature maps. As the second attention mechanism, compatibility scores across spatial locations and output of our DNN highlights the most discriminative regions, and lastly, the multiheaded self-attention augmented convolutions account for our third attention mechanism. We evaluate the efficiency of our proposed framework through extensive experiments using multiple morph datasets that are compiled using bona fide images available in the FERET, FRLL, FRGC, and WVU Twin datasets. Most importantly, our proposed methodology has resulted in reduction in detection error rates when compared with state-of-the-art results. Finally, to further assess our multi-attentional morph detection, we delve into different combinations of attention mechanisms via a comprehensive ablation study.
- 9 March 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 24281-24298
Open AccessAttention Augmented Face Morph Detection
Morph detection is of paramount significance when the integrity of Automatic Face Recognition (AFR) systems is concerned. Considering the risks incurred by morphing attacks, a robust automated morph detector is required which can distinguish authentic bona fide samples from altered morphed images. We leverage the wavelet sub-band decomposition of an input image, yielding the fine-grained spatial-frequency content of the input image. To enhance detection of morphed images, our goal is to find the most discriminative information across frequency channels and spatial domain. To this end, we propose an end-to-end attention-based deep morph detector which assimilates the most discriminative wavelet sub-bands of a given image which are obtained by a group sparsity representation learning scheme. Specifically, our group sparsity-constrained Deep Neural Network (DNN) learns the most discriminative wavelet sub-bands (channels) of an input image while the attention mechanism captures the most discriminative spatial regions of input images for the downstream task of morph detection. To this end, we adopt three attention mechanisms to diversify our refined features for morph detection. As the first attention mechanism, we employ the Convolutional Block Attention Module (CBAM) which provides us with refined feature maps. As the second attention mechanism, compatibility scores across spatial locations and output of our DNN highlights the most discriminative regions, and lastly, the multiheaded self-attention augmented convolutions account for our third attention mechanism. We evaluate the efficiency of our proposed framework through extensive experiments using multiple morph datasets that are compiled using bona fide images available in the FERET, FRLL, FRGC, and WVU Twin datasets. Most importantly, our proposed methodology has resulted in reduction in detection error rates when compared with state-of-the-art results. Finally, to further assess our multi-attentional morph detection, we delve into different combinations of attention mechanisms via a comprehensive ablation study.
- Open Access
- 30 August 2022
- Published by Institution of Engineering and Technology (IET) in IET Biometrics
- Vol. 11 (5) , 459-484
Benchmarking human face similarity using identical twins
The problem of distinguishing identical twins and non‐twin look‐alikes in automated facial recognition (FR) applications has become increasingly important with the widespread adoption of facial biometrics. Due to the high facial similarity of both identical twins and look‐alikes, these face pairs represent the hardest cases presented to facial recognition tools. This work presents an application of one of the largest twin data sets compiled to date to address two FR challenges: (1) determining a baseline measure of facial similarity between identical twins and (2) applying this similarity measure to determine the impact of doppelgangers, or look‐alikes, on FR performance for large face data sets. The facial similarity measure is determined via a deep convolutional neural network. This network is trained on a tailored verification task designed to encourage the network to group together highly similar face pairs in the embedding space and achieves a test AUC of 0.9799. The proposed network provides a quantitative similarity score for any two given faces and has been applied to large‐scale face data sets to identify similar face pairs. An additional analysis that correlates the comparison score returned by a facial recognition tool and the similarity score returned by the proposed network has also been performed.
- 30 August 2022
- Published by Institution of Engineering and Technology (IET) in IET Biometrics
- Vol. 11 (5) , 459-484
Open AccessBenchmarking human face similarity using identical twins
The problem of distinguishing identical twins and non‐twin look‐alikes in automated facial recognition (FR) applications has become increasingly important with the widespread adoption of facial biometrics. Due to the high facial similarity of both identical twins and look‐alikes, these face pairs represent the hardest cases presented to facial recognition tools. This work presents an application of one of the largest twin data sets compiled to date to address two FR challenges: (1) determining a baseline measure of facial similarity between identical twins and (2) applying this similarity measure to determine the impact of doppelgangers, or look‐alikes, on FR performance for large face data sets. The facial similarity measure is determined via a deep convolutional neural network. This network is trained on a tailored verification task designed to encourage the network to group together highly similar face pairs in the embedding space and achieves a test AUC of 0.9799. The proposed network provides a quantitative similarity score for any two given faces and has been applied to large‐scale face data sets to identify similar face pairs. An additional analysis that correlates the comparison score returned by a facial recognition tool and the similarity score returned by the proposed network has also been performed.