Search

This work explores how human judgement about salient regions of an image can be introduced into deep convolutional neural network (DCNN) training. Traditionally, training of DCNNs is purely data-driven. This often results in learning features of the data that are only coincidentally correlated with class labels. Human saliency can guide network training using our proposed new component of the loss function that ConveYs Brain Oversight to Raise Generalization (CYBORG) and penalizes the model for using non-salient regions. This mechanism produces DCNNs achieving higher accuracy and generalization compared to using the same training data without human salience. Experimental results demonstrate that CYBORG applies across multiple network architectures and problem domains (detection of synthetic faces, iris presentation attacks and anomalies in chest X-rays), while requiring significantly less data than training without human saliency guidance. Visualizations show that CYBORG-trained models' saliency is more consistent across independent training runs than traditionally-trained models, and also in better agreement with humans. To lower the cost of collecting human annotations, we also explore using deep learning to provide automated annotations. CYBORG training of CNNs addresses important issues such as reducing the appetite for large training sets, increasing interpretability, and reducing fragility by generalizing better to new types of data.

This paper studies how to synthesize face images of non-existent persons, to create a dataset that allows effective training of face recognition (FR) models. Two important goals are (1) the ability to generate a large number of distinct identities (inter-class separation) with (2) a wide variation in appearance of each identity (intra-class variation). However, existing works 1) are typically limited in how many well-separated identities can be generated and 2) either neglect or use a separate editing model for attribute augmentation. We propose Vec2Face, a holistic model that uses only a sampled vector as input and can flexibly generate and control face images and their attributes. Composed of a feature masked autoencoder and a decoder, Vec2Face is supervised by face image reconstruction and can be conveniently used in inference. Using vectors with low similarity among themselves as inputs, Vec2Face generates well-separated identities. Randomly perturbing an input identity vector within a small range allows Vec2Face to generate faces of the same identity with robust variation in face attributes. It is also possible to generate images with designated attributes by adjusting vector values with a gradient descent method. Vec2Face has efficiently synthesized as many as 300K identities with 15 million total images, whereas 60K is the largest number of identities created in the previous works. FR models trained with the generated HSFace datasets, from 10k to 300k identities, achieve state-of-the-art accuracy, from 92% to 93.52%, on five real-world test sets. For the first time, our model created using a synthetic training set achieves higher accuracy than the model created using a same-scale training set of real face images (on the CALFW test set).

Appearance of a face can be greatly altered by growing a beard and mustache. The facial hairstyles in a pair of images can cause marked changes to the impostor distribution and the genuine distribution. Also, different distributions of facial hairstyle across demographics could cause a false impression of relative accuracy across demographics. We first show that, even though larger training sets boost the recognition accuracy on all facial hairstyles, accuracy variations caused by facial hairstyles persist regardless of the size of the training set. Then, we analyze the impact of having different fractions of the training data represent facial hairstyles. We created balanced training sets using a set of identities available in Webface42M that both have clean-shaven and facial hair images. We find that, even when a face recognition model is trained with a balanced clean-shaven / facial hair training set, accuracy variation on the test data does not diminish. Next, data augmentation is employed to further investigate the effect of facial hair distribution in training data by manipulating facial hair pixels with the help of facial landmark points and a facial hair segmentation model. Our results show facial hair causes an accuracy gap between clean-shaven and facial hair images, and this impact can be significantly different between African-Americans and Caucasians.

Face Recognition models are commonly trained with web-scraped datasets containing millions of images and evaluated on test sets emphasizing pose, age and mixed attributes. With train and test sets both assembled from web-scraped images, it is critical to ensure disjoint sets of identities between train and test sets. However, existing train and test sets have not considered this. Moreover, as accuracy levels become saturated, such as LFW $>99.8\%$, more challenging test sets are needed. We show that current train and test sets are generally not identity- or even image-disjoint, and that this results in an optimistic bias in the estimated accuracy. In addition, we show that identity-disjoint folds are important in the 10-fold cross-validation estimate of test accuracy. To better support continued advances in face recognition, we introduce two "Goldilocks" test sets, Hadrian and Eclipse. The former emphasizes challenging facial hairstyles and latter emphasizes challenging over- and under-exposure conditions. Images in both datasets are from a large, controlled-acquisition (not web-scraped) dataset, so they are identity- and image-disjoint with all popular training sets. Accuracy for these new test sets generally falls below that observed on LFW, CPLFW, CALFW, CFP-FP and AgeDB-30, showing that these datasets represent important dimensions for improvement of face recognition. The datasets are available at: \url{https://github.com/HaiyuWu/SOTA-Face-Recognition-Train-and-Test}

A fundamental tenet of pattern recognition is that overlap between training and testing sets causes an optimistic accuracy estimate. Deep CNNs for face recognition are trained for N-way classification of the identities in the training set. Accuracy is commonly estimated as average 10-fold classification accuracy on image pairs from test sets such as LFW, CALFW, CPLFW, CFP-FP and AgeDB-30. Because train and test sets have been independently assembled, images and identities in any given test set may also be present in any given training set. In particular, our experiments reveal a surprising degree of identity and image overlap between the LFW family of test sets and the MS1MV2 training set. Our experiments also reveal identity label noise in MS1MV2. We compare accuracy achieved with same-size MS1MV2 subsets that are identity-disjoint and not identity-disjoint with LFW, to reveal the size of the optimistic bias. Using more challenging test sets from the LFW family, we find that the size of the optimistic bias is larger for more challenging test sets. Our results highlight the lack of and the need for identity-disjoint train and test methodology in face recognition research.

Synthesis of same-identity biometric iris images, both for existing and non-existing identities while preserving the identity across a wide range of pupil sizes, is complex due to intricate iris muscle constriction mechanism, requiring a precise model of iris non-linear texture deformations to be embedded into the synthesis pipeline. This paper presents the first method of fully data-driven, identity-preserving, pupil size-varying s ynthesis of iris images. This approach is capable of synthesizing images of irises with different pupil sizes representing non-existing identities as well as non-linearly deforming the texture of iris images of existing subjects given the segmentation mask of the target iris image. Iris recognition experiments suggest that the proposed deformation model not only preserves the identity when changing the pupil size but offers better similarity between same-identity iris samples with significant differences in pupil size, compared to state-of-the-art linear and non-linear (bio-mechanical-based) iris deformation models. Two immediate applications of the proposed approach are: (a) synthesis of, or enhancement of the existing biometric datasets for iris recognition, mimicking those acquired with iris sensors, and (b) helping forensic human experts in examining iris image pairs with significant differences in pupil dilation. Source codes and weights of the models are made available with the paper.

The first layer of a deep CNN backbone applies filters to an image to extract the basic features available to later layers. During training, some filters may go inactive, mean ing all weights in the filter approach zero. An inactive fil ter in the final model represents a missed opportunity to extract a useful feature. This phenomenon is especially prevalent in specialized CNNs such as for face recogni tion (as opposed to, e.g., ImageNet). For example, in one the most widely face recognition model (ArcFace), about half of the convolution filters in the first layer are inactive. We propose a novel approach designed and tested specif ically for face recognition networks, known as "CRAFT: Contextual Re-Activation of Filters for Face Recognition Training". CRAFT identifies inactive filters during training and reinitializes them based on the context of strong filters at that stage in training. We show that CRAFT reduces fraction of inactive filters from 44% to 32% on average and discovers filter patterns not found by standard training. Compared to standard training without reactivation, CRAFT demonstrates enhanced model accuracy on standard face-recognition benchmark datasets including AgeDB-30, CPLFW, LFW, CALFW, and CFP-FP, as well as on more challenging datasets like IJBB and IJBC.

Ensuring logical consistency in predictions is a crucial yet overlooked aspect in multi-attribute classification. We explore the potential reasons for this oversight and introduce two pressing challenges to the field: 1) How can we ensure that a model, when trained with data checked for logical consistency, yields predictions that are logically consistent? 2) How can we achieve the same with data that hasn't undergone logical consistency checks? Minimizing manual effort is also essential for enhancing automation. To address these challenges, we introduce two datasets, FH41K and CelebA-logic, and propose LogicNet, an adversarial training framework that learns the logical relationships between attributes. Accuracy of LogicNet surpasses that of the next-best approach by 23.05%, 9.96%, and 1.71% on FH37K, FH41K, and CelebA-logic, respectively. In real-world case analysis, our approach can achieve a reduction of more than 50% in the average number of failed cases compared to other methods.

State-of-the-art deep CNN face matchers are typically created using extensive training sets of color face images. Our study reveals that such matchers attain virtually identical accuracy when trained on either grayscale or color versions of the training set, even when the evaluation is done using color test images. Furthermore, we demonstrate that shallower models, lacking the capacity to model complex representations, rely more heavily on low-level features such as those associated with color. As a result, they display diminished accuracy when trained with grayscale images. We then consider possible causes for deeper CNN face matchers "not seeing color". Popular web-scraped face datasets actually have 30 to 60% of their identities with one or more grayscale images. We analyze whether this grayscale element in the training set impacts the accuracy achieved, and conclude that it does not. We demonstrate that using only grayscale images for both training and testing achieves accuracy comparable to that achieved using only color images for deeper models. This holds true for both real and synthetic training datasets. HSV color space, which separates chroma and luma information, does not improve the network's learning about color any more than in the RGB color space. We then show that the skin region of an individual's images in a web-scraped training set exhibits significant variation in their mapping to color space. This suggests that color carries limited identity-specific information. We also show that when the first convolution layer is restricted to a single filter, models learn a grayscale conversion filter and pass a grayscale version of the input color image to the next layer. Finally, we demonstrate that leveraging the lower per-image storage for grayscale to increase the number of images in the training set can improve accuracy of the face recognition model., Comment: This is replacement version of the previous arxiv submission: 2309.05180 (Our Deep CNN Face Matchers Have Developed Achromatopsia). The past version is published in CVPRW and available in IEEE proceedings. This submitted version is an extension of the conference paper

Most studies to date that have examined demographic variations in face recognition accuracy have analyzed 1-to-1 matching accuracy, using images that could be described as "government ID quality". This paper analyzes the accuracy of 1-to-many facial identification across demographic groups, and in the presence of blur and reduced resolution in the probe image as might occur in "surveillance camera quality" images. Cumulative match characteristic curves (CMC) are not appropriate for comparing propensity for rank-one recognition errors across demographics, and so we use three metrics for our analysis: (1) the well-known d' metric between mated and non-mated score distributions, and introduced in this work, (2) absolute score difference between thresholds in the high-similarity tail of the non-mated and the low-similarity tail of the mated distribution, and (3) distribution of (mated - non-mated rank-one scores) across the set of probe images. We find that demographic variation in 1-to-many accuracy does not entirely follow what has been observed in 1-to-1 matching accuracy. Also, different from 1-to-1 accuracy, demographic comparison of 1-to-many accuracy can be affected by different numbers of identities and images across demographics. More importantly, we show that increased blur in the probe image, or reduced resolution of the face in the probe image, can significantly increase the false positive identification rate. And we show that the demographic variation in these high blur or low resolution conditions is much larger for male / female than for African-American / Caucasian. The point that 1-to-many accuracy can potentially collapse in the context of processing "surveillance camera quality" probe images against a "government ID quality" gallery is an important one., Comment: 9 pages, 8 figures, Conference submission

The issue of demographic disparities in face recognition accuracy has attracted increasing attention in recent years. Various face image datasets have been proposed as 'fair' or 'balanced' to assess the accuracy of face recognition algorithms across demographics. These datasets typically balance the number of identities and images across demographics. It is important to note that the number of identities and images in an evaluation dataset are {\em not} driving factors for 1-to-1 face matching accuracy. Moreover, balancing the number of identities and images does not ensure balance in other factors known to impact accuracy, such as head pose, brightness, and image quality. We demonstrate these issues using several recently proposed datasets. To improve the ability to perform less biased evaluations, we propose a bias-aware toolkit that facilitates creation of cross-demographic evaluation datasets balanced on factors mentioned in this paper.

In recent years, media reports have called out bias and racism in face recognition technology. We review experimental results exploring several speculated causes for asymmetric cross-demographic performance. We consider accuracy differences as represented by variations in non-mated (impostor) and / or mated (genuine) distributions for 1-to-1 face matching. Possible causes explored include differences in skin tone, face size and shape, imbalance in number of identities and images in the training data, and amount of face visible in the test data ("face pixels"). We find that demographic differences in face pixel information of the test images appear to most directly impact the resultant differences in face recognition accuracy.

Face attribute research has so far used only simple binary attributes for facial hair; e.g., beard / no beard. We have created a new, more descriptive facial hair annotation scheme and applied it to create a new facial hair attribute dataset, FH37K. Face attribute research also so far has not dealt with logical consistency and completeness. For example, in prior research, an image might be classified as both having no beard and also having a goatee (a type of beard). We show that the test accuracy of previous classification methods on facial hair attribute classification drops significantly if logical consistency of classifications is enforced. We propose a logically consistent prediction loss, LCPLoss, to aid learning of logical consistency across attributes, and also a label compensation training strategy to eliminate the problem of no positive prediction across a set of related attributes. Using an attribute classifier trained on FH37K, we investigate how facial hair affects face recognition accuracy, including variation across demographics. Results show that similarity and difference in facial hairstyle have important effects on the impostor and genuine score distributions in face recognition. The code is at https:// github.com/ HaiyuWu/ LogicalConsistency.

We report the first systematic analysis of the experimental foundations of facial attribute classification. Two annotators independently assigning attribute values shows that only 12 of 40 common attributes are assigned values with >= 95% consistency, and three (high cheekbones, pointed nose, oval face) have essentially random consistency. Of 5,068 duplicate face appearances in CelebA, attributes have contradicting values on from 10 to 860 of the 5,068 duplicates. Manual audit of a subset of CelebA estimates error rates as high as 40% for (no beard=false), even though the labeling consistency experiment indicates that no beard could be assigned with >= 95% consistency. Selecting the mouth slightly open (MSO) for deeper analysis, we estimate the error rate for (MSO=true) at about 20% and (MSO=false) at about 2%. A corrected version of the MSO attribute values enables learning a model that achieves higher accuracy than previously reported for MSO. Corrected values for CelebA MSO are available at https://github.com/HaiyuWu/CelebAMSO.

It is broadly accepted that there is a "gender gap" in face recognition accuracy, with females having higher false match and false non-match rates. However, relatively little is known about the cause(s) of this gender gap. Even the recent NIST report on demographic effects lists "analyze cause and effect" under "what we did not do". We first demonstrate that female and male hairstyles have important differences that impact face recognition accuracy. In particular, compared to females, male facial hair contributes to creating a greater average difference in appearance between different male faces. We then demonstrate that when the data used to estimate recognition accuracy is balanced across gender for how hairstyles occlude the face, the initially observed gender gap in accuracy largely disappears. We show this result for two different matchers, and analyzing images of Caucasians and of African-Americans. These results suggest that future research on demographic variation in accuracy should include a check for balanced quality of the test data as part of the problem formulation. To promote reproducible research, matchers, attribute classifiers, and datasets used in this research are/will be publicly available.

We explore varying face recognition accuracy across demographic groups as a phenomenon partly caused by differences in face illumination. We observe that for a common operational scenario with controlled image acquisition, there is a large difference in face region brightness between African-American and Caucasian, and also a smaller difference between male and female. We show that impostor image pairs with both faces under-exposed, or both overexposed, have an increased false match rate (FMR). Conversely, image pairs with strongly different face brightness have a decreased similarity measure. We propose a brightness information metric to measure variation in brightness in the face and show that face brightness that is too low or too high has reduced information in the face region, providing a cause for the lower accuracy. Based on this, for operational scenarios with controlled image acquisition, illumination should be adjusted for each individual to obtain appropriate face image brightness. This is the first work that we are aware of to explore how the level of brightness of the skin region in a pair of face images (rather than a single image) impacts face recognition accuracy, and to evaluate this as a systematic factor causing unequal accuracy across demographics. The code is at https://github.com/HaiyuWu/FaceBrightness.

Media reports have accused face recognition of being ''biased'', ''sexist'' and ''racist''. There is consensus in the research literature that face recognition accuracy is lower for females, who often have both a higher false match rate and a higher false non-match rate. However, there is little published research aimed at identifying the cause of lower accuracy for females. For instance, the 2019 Face Recognition Vendor Test that documents lower female accuracy across a broad range of algorithms and datasets also lists ''Analyze cause and effect'' under the heading ''What we did not do''. We present the first experimental analysis to identify major causes of lower face recognition accuracy for females on datasets where previous research has observed this result. Controlling for equal amount of visible face in the test images mitigates the apparent higher false non-match rate for females. Additional analysis shows that makeup-balanced datasets further improves females to achieve lower false non-match rates. Finally, a clustering experiment suggests that images of two different females are inherently more similar than of two different males, potentially accounting for a difference in false match rates., Comment: arXiv admin note: substantial text overlap with arXiv:2008.06989

Remote photoplethysmography (rPPG) is a technique for estimating blood volume changes from reflected light without the need for a contact sensor. We present the first examples of presentation attacks in the digital and physical domains on rPPG from face video. Digital attacks are easily performed by adding imperceptible periodic noise to the input videos. Physical attacks are performed with illumination from visible spectrum LEDs placed in close proximity to the face, while still being difficult to perceive with the human eye. We also show that our attacks extend beyond medical applications, since the method can effectively generate a strong periodic pulse on 3D-printed face masks, which presents difficulties for pulse-based face presentation attack detection (PAD). The paper concludes with ideas for using this work to improve robustness of rPPG methods and pulse-based face PAD.

We present the Deception Detection and Physiological Monitoring (DDPM) dataset and initial baseline results on this dataset. Our application context is an interview scenario in which the interviewee attempts to deceive the interviewer on selected responses. The interviewee is recorded in RGB, near-infrared, and long-wave infrared, along with cardiac pulse, blood oxygenation, and audio. After collection, data were annotated for interviewer/interviewee, curated, ground-truthed, and organized into train / test parts for a set of canonical deception detection experiments. Baseline experiments found random accuracy for micro-expressions as an indicator of deception, but that saccades can give a statistically significant response. We also estimated subject heart rates from face videos (remotely) with a mean absolute error as low as 3.16 bpm. The database contains almost 13 hours of recordings of 70 subjects, and over 8 million visible-light, near-infrared, and thermal video frames, along with appropriate meta, audio and pulse oximeter data. To our knowledge, this is the only collection offering recordings of five modalities in an interview scenario that can be used in both deception detection and remote photoplethysmography research., Comment: The dataset will be available for download at https://cvrl.nd.edu/projects/data/#deception-detection-and-physiological-monitoringddpm

News reports have suggested that darker skin tone causes an increase in face recognition errors. The Fitzpatrick scale is widely used in dermatology to classify sensitivity to sun exposure and skin tone. In this paper, we analyze a set of manual Fitzpatrick skin type assignments and also employ the individual typology angle to automatically estimate the skin tone from face images. The set of manual skin tone rating experiments shows that there are inconsistencies between human raters that are difficult to eliminate. Efforts to automate skin tone rating suggest that it is particularly challenging on images collected without a calibration object in the scene. However, after the color-correction, the level of agreement between automated and manual approaches is found to be 96% or better for the MORPH images. To our knowledge, this is the first work to: (a) examine the consistency of manual skin tone ratings across observers, (b) document that there is substantial variation in the rating of the same image by different observers even when exemplar images are given for guidance and all images are color-corrected, and (c) compare manual versus automated skin tone ratings.

This paper is the first to explore the question of whether images that are classified incorrectly by a face analytics algorithm (e.g., gender classification) are any more or less likely to participate in an image pair that results in a face recognition error. We analyze results from three different gender classification algorithms (one open-source and two commercial), and two face recognition algorithms (one open-source and one commercial), on image sets representing four demographic groups (African-American female and male, Caucasian female and male). For impostor image pairs, our results show that pairs in which one image has a gender classification error have a better impostor distribution than pairs in which both images have correct gender classification, and so are less likely to generate a false match error. For genuine image pairs, our results show that individuals whose images have a mix of correct and incorrect gender classification have a worse genuine distribution (increased false non-match rate) compared to individuals whose images all have correct gender classification. Thus, compared to images that generate correct gender classification, images that generate gender classification errors do generate a different pattern of recognition errors, both better (false match) and worse (false non-match).

Recent news articles have accused face recognition of being "biased", "sexist" or "racist". There is consensus in the research literature that face recognition accuracy is lower for females, who often have both a higher false match rate and a higher false non-match rate. However, there is little published research aimed at identifying the cause of lower accuracy for females. For instance, the 2019 Face Recognition Vendor Test that documents lower female accuracy across a broad range of algorithms and datasets also lists "Analyze cause and effect" under the heading "What we did not do". We present the first experimental analysis to identify major causes of lower face recognition accuracy for females on datasets where previous research has observed this result. Controlling for equal amount of visible face in the test images reverses the apparent higher false non-match rate for females. Also, principal component analysis indicates that images of two different females are inherently more similar than of two different males, potentially accounting for a difference in false match rates., Comment: Paper will appear on BMVC 2020

As the popularity of iris recognition systems increases, the importance of effective security measures against presentation attacks becomes paramount. This work presents an overview of the most important advances in the area of iris presentation attack detection published in recent two years. Newly-released, publicly-available datasets for development and evaluation of iris presentation attack detection are discussed. Recent literature can be seen to be broken into three categories: traditional "hand-crafted" feature extraction and classification, deep learning-based solutions, and hybrid approaches fusing both methodologies. Conclusions of modern approaches underscore the difficulty of this task. Finally, commentary on possible directions for future research is provided., Comment: Under revision for Pattern Recognition Letters

The automatic analysis of face images can generate predictions about a person's gender, age, race, facial expression, body mass index, and various other indices and conditions. A few recent publications have claimed success in analyzing an image of a person's face in order to predict the person's status as Criminal / Non-Criminal. Predicting criminality from face may initially seem similar to other facial analytics, but we argue that attempts to create a criminality-from-face algorithm are necessarily doomed to fail, that apparently promising experimental results in recent publications are an illusion resulting from inadequate experimental design, and that there is potentially a large social cost to belief in the criminality from face illusion.

Web-scraped, in-the-wild datasets have become the norm in face recognition research. The numbers of subjects and images acquired in web-scraped datasets are usually very large, with number of images on the millions scale. A variety of issues occur when collecting a dataset in-the-wild, including images with the wrong identity label, duplicate images, duplicate subjects and variation in quality. With the number of images being in the millions, a manual cleaning procedure is not feasible. But fully automated methods used to date result in a less-than-ideal level of clean dataset. We propose a semi-automated method, where the goal is to have a clean dataset for testing face recognition methods, with similar quality across men and women, to support comparison of accuracy across gender. Our approach removes near-duplicate images, merges duplicate subjects, corrects mislabeled images, and removes images outside a defined range of pose and quality. We conduct the curation on the Asian Face Dataset (AFD) and VGGFace2 test dataset. The experiments show that a state-of-the-art method achieves a much higher accuracy on the datasets after they are curated. Finally, we release our cleaned versions of both datasets to the research community., Comment: This paper will appear at IWBF 2020

Diversity and unpredictability of artifacts potentially presented to an iris sensor calls for presentation attack detection methods that are agnostic to specificity of presentation attack instruments. This paper proposes a method that combines two-dimensional and three-dimensional properties of the observed iris to address the problem of spoof detection in case when some properties of artifacts are unknown. The 2D (textural) iris features are extracted by a state-of-the-art method employing Binary Statistical Image Features (BSIF) and an ensemble of classifiers is used to deliver 2D modality-related decision. The 3D (shape) iris features are reconstructed by a photometric stereo method from only two images captured under near-infrared illumination placed at two different angles, as in many current commercial iris recognition sensors. The map of normal vectors is used to assess the convexity of the observed iris surface. The combination of these two approaches has been applied to detect whether a subject is wearing a textured contact lens to disguise their identity. Extensive experiments with NDCLD'15 dataset, and a newly collected NDIris3D dataset show that the proposed method is highly robust under various open-set testing scenarios, and that it outperforms all available open-source iris PAD methods tested in identical scenarios. The source code and the newly prepared benchmark are made available along with this paper., Comment: Accepted to IEEE T-IFS

Deep learning methods have greatly increased the accuracy of face recognition, but an old problem still persists: accuracy is usually higher for men than women. It is often speculated that lower accuracy for women is caused by under-representation in the training data. This work investigates female under-representation in the training data is truly the cause of lower accuracy for females on test data. Using a state-of-the-art deep CNN, three different loss functions, and two training datasets, we train each on seven subsets with different male/female ratios, totaling forty two trainings, that are tested on three different datasets. Results show that (1) gender balance in the training data does not translate into gender balance in the test accuracy, (2) the "gender gap" in test accuracy is not minimized by a gender-balanced training set, but by a training set with more male images than female images, and (3) training to minimize the accuracy gap does not result in highest female, male or average accuracy

We present a comprehensive analysis of how and why face recognition accuracy differs between men and women. We show that accuracy is lower for women due to the combination of (1) the impostor distribution for women having a skew toward higher similarity scores, and (2) the genuine distribution for women having a skew toward lower similarity scores. We show that this phenomenon of the impostor and genuine distributions for women shifting closer towards each other is general across datasets of African-American, Caucasian, and Asian faces. We show that the distribution of facial expressions may differ between male/female, but that the accuracy difference persists for image subsets rated confidently as neutral expression. The accuracy difference also persists for image subsets rated as close to zero pitch angle. Even when removing images with forehead partially occluded by hair/hat, the same impostor/genuine accuracy difference persists. We show that the female genuine distribution improves when only female images without facial cosmetics are used, but that the female impostor distribution also degrades at the same time. Lastly, we show that the accuracy difference persists even if a state-of-the-art deep learning method is trained from scratch using training data explicitly balanced between male and female images and subjects., Comment: Paper will appear at The 2nd Workshop on Demographic Variation in the Performance of Biometric Systems at WACV 2020

Matching live images (``selfies'') to images from ID documents is a problem that can arise in various applications. A challenging instance of the problem arises when the face image on the ID document is from early adolescence and the live image is from later adolescence. We explore this problem using a private dataset called Chilean Young Adult (CHIYA) dataset, where we match live face images taken at age 18-19 to face images on ID documents created at ages 9 to 18. State-of-the-art deep learning face matchers (e.g., ArcFace) have relatively poor accuracy for document-to-selfie face matching. To achieve higher accuracy, we fine-tune the best available open-source model with triplet loss for a few-shot learning. Experiments show that our approach achieves higher accuracy than the DocFace+ model recently developed for this problem. Our fine-tuned model was able to improve the true acceptance rate for the most difficult (largest age span) subset from 62.92% to 96.67% at a false acceptance rate of 0.01%. Our fine-tuned model is available for use by other researchers.

Previous studies generally agree that face recognition accuracy is higher for older persons than for younger persons. But most previous studies were before the wave of deep learning matchers, and most considered accuracy only in terms of the verification rate for genuine pairs. This paper investigates accuracy for age groups 16-29, 30-49 and 50-70, using three modern deep CNN matchers, and considers differences in the impostor and genuine distributions as well as verification rates and ROC curves. We find that accuracy is lower for older persons and higher for younger persons. In contrast, a pre deep learning matcher on the same dataset shows the traditional result of higher accuracy for older persons, although its overall accuracy is much lower than that of the deep learning matchers. Comparing the impostor and genuine distributions, we conclude that impostor scores have a larger effect than genuine scores in causing lower accuracy for the older age group. We also investigate the effects of training data across the age groups. Our results show that fine-tuning the deep CNN models on additional images of older persons actually lowers accuracy for the older age group. Also, we fine-tune and train from scratch two models using age-balanced training datasets, and these results also show lower accuracy for older age group. These results argue that the lower accuracy for the older age group is not due to imbalance in the original training data., Comment: Paper will appear at the WACV 2020

We propose a new iris presentation attack detection method using three-dimensional features of an observed iris region estimated by photometric stereo. Our implementation uses a pair of iris images acquired by a common commercial iris sensor (LG 4000). No hardware modifications of any kind are required. Our approach should be applicable to any iris sensor that can illuminate the eye from two different directions. Each iris image in the pair is captured under near-infrared illumination at a different angle relative to the eye. Photometric stereo is used to estimate surface normal vectors in the non-occluded portions of the iris region. The variability of the normal vectors is used as the presentation attack detection score. This score is larger for a texture that is irregularly opaque and printed on a convex contact lens, and is smaller for an authentic iris texture. Thus the problem is formulated as binary classification into (a) an eye wearing textured contact lens and (b) the texture of an actual iris surface (possibly seen through a clear contact lens). Experiments were carried out on a database of approx. 2,900 iris image pairs acquired from approx. 100 subjects. Our method was able to correctly classify over 95% of samples when tested on contact lens brands unseen in training, and over 98% of samples when the contact lens brand was seen during training. The source codes of the method are made available to other researchers., Comment: Patent Pending. Paper accepted for WACV 2019, Hawaii, USA

We propose a multi-scale GAN model to hallucinate realistic context (forehead, hair, neck, clothes) and background pixels automatically from a single input face mask. Instead of swapping a face on to an existing picture, our model directly generates realistic context and background pixels based on the features of the provided face mask. Unlike face inpainting algorithms, it can generate realistic hallucinations even for a large number of missing pixels. Our model is composed of a cascaded network of GAN blocks, each tasked with hallucination of missing pixels at a particular resolution while guiding the synthesis process of the next GAN block. The hallucinated full face image is made photo-realistic by using a combination of reconstruction, perceptual, adversarial and identity preserving losses at each block of the network. With a set of extensive experiments, we demonstrate the effectiveness of our model in hallucinating context and background pixels from face masks varying in facial pose, expression and lighting, collected from multiple datasets subject disjoint with our training data. We also compare our method with two popular face swapping and face completion methods in terms of visual quality and recognition performance. Additionally, we analyze our cascaded pipeline and compare it with the recently proposed progressive growing of GANs., Comment: Extended version of WACV 2020 paper

We propose an algorithm to generate realistic face images of both real and synthetic identities (people who do not exist) with different facial yaw, shape and resolution.The synthesized images can be used to augment datasets to train CNNs or as massive distractor sets for biometric verification experiments without any privacy concerns. Additionally, law enforcement can make use of this technique to train forensic experts to recognize faces. Our method samples face components from a pool of multiple face images of real identities to generate the synthetic texture. Then, a real 3D head model compatible to the generated texture is used to render it under different facial yaw transformations. We perform multiple quantitative experiments to assess the effectiveness of our synthesis procedure in CNN training and its potential use to generate distractor face images. Additionally, we compare our method with popular GAN models in terms of visual quality and execution time., Comment: To appear in IEEE WACV 2019. Get our data (2M face images of 12K synthetic subjects, 8K 3D head models) by accessing the "Notre Dame Synthetic Face Dataset" here: https://cvrl.nd.edu/projects/data/

This paper proposes the first, known to us, open source presentation attack detection (PAD) solution to distinguish between authentic iris images (possibly wearing clear contact lenses) and irises with textured contact lenses. This software can serve as a baseline in various PAD evaluations, and also as an open-source platform with an up-to-date reference method for iris PAD. The software is written in C++ and Python and uses only open source resources, such as OpenCV. This method does not incorporate iris image segmentation, which may be problematic for unknown fake samples. Instead, it makes a best guess to localize the rough position of the iris. The PAD-related features are extracted with the Binary Statistical Image Features (BSIF), which are classified by an ensemble of classifiers incorporating support vector machine, random forest and multilayer perceptron. The models attached to the current software have been trained with the NDCLD'15 database and evaluated on the independent datasets included in the LivDet-Iris 2017 competition. The software implements the functionality of retraining the classifiers with any database of authentic and attack images. The accuracy of the current version offered with this paper exceeds 99% when tested on subject-disjoint subsets of NDCLD'15, and oscillates around 85% when tested on the LivDet-Iris 2017 benchmarks, which is on par with the results obtained by the LivDet-Iris 2017 winner.

Binarized statistical image features (BSIF) have been successfully used for texture analysis in many computer vision tasks, including iris recognition and biometric presentation attack detection. One important point is that all applications of BSIF in iris recognition have used the original BSIF filters, which were trained on image patches extracted from natural images. This paper tests the question of whether domain-specific BSIF can give better performance than the default BSIF. The second important point is in the selection of image patches to use in training for BSIF. Can image patches derived from eye-tracking experiments, in which humans perform an iris recognition task, give better performance than random patches? Our results say that (1) domain-specific BSIF features can out-perform the default BSIF features, and (2) selecting image patches in a task-specific manner guided by human performance can out-perform selecting random patches. These results are important because BSIF is often regarded as a generic texture tool that does not need any domain adaptation, and human-task-guided selection of patches for training has never (to our knowledge) been done. This paper follows the reproducible research requirements, and the new iris-domain-specific BSIF filters, the patches used in filter training, the database used in testing and the source codes of the designed iris recognition method are made available along with this paper to facilitate applications of this concept., Comment: Paper accepted for WACV 2019, Hawaii, USA

This paper advances the state of the art in human examination of iris images by (1) assessing the impact of different iris conditions in identity verification, and (2) introducing an annotation step that improves the accuracy of people's decisions. In a first experimental session, 114 subjects were asked to decide if pairs of iris images depict the same eye (genuine pairs) or two distinct eyes (impostor pairs). The image pairs sampled six conditions: (1) easy for algorithms to classify, (2) difficult for algorithms to classify, (3) large difference in pupil dilation, (4) disease-affected eyes, (5) identical twins, and (6) post-mortem samples. In a second session, 85 of the 114 subjects were asked to annotate matching and non-matching regions that supported their decisions. Subjects were allowed to change their initial classification as a result of the annotation process. Results suggest that: (a) people improve their identity verification accuracy when asked to annotate matching and non-matching regions between the pair of images, (b) images depicting the same eye with large difference in pupil dilation were the most challenging to subjects, but benefited well from the annotation-driven classification, (c) humans performed better than iris recognition algorithms when verifying genuine pairs of post-mortem and disease-affected eyes (i.e., samples showing deformations that go beyond the distortions of a healthy iris due to pupil dilation), and (d) annotation does not improve accuracy of analyzing images from identical twins, which remain confusing for people., Comment: Paper accepted for WACV 2019, Hawaii, USA

Creative works, whether paintings or memes, follow unique journeys that result in their final form. Understanding these journeys, a process known as "provenance analysis", provides rich insights into the use, motivation, and authenticity underlying any given work. The application of this type of study to the expanse of unregulated content on the Internet is what we consider in this paper. Provenance analysis provides a snapshot of the chronology and validity of content as it is uploaded, re-uploaded, and modified over time. Although still in its infancy, automated provenance analysis for online multimedia is already being applied to different types of content. Most current works seek to build provenance graphs based on the shared content between images or videos. This can be a computationally expensive task, especially when considering the vast influx of content that the Internet sees every day. Utilizing non-content-based information, such as timestamps, geotags, and camera IDs can help provide important insights into the path a particular image or video has traveled during its time on the Internet without large computational overhead. This paper tests the scope and applicability of metadata-based inferences for provenance graph construction in two different scenarios: digital image forensics and cultural analytics., Comment: Supplemental material for this paper can be found at https://drive.google.com/file/d/1Tbs2CQg_VQAc2PdztW5twVaiXD0G12-H/view?usp=sharing

Iris recognition is increasingly used in large-scale applications. As a result, presentation attack detection for iris recognition takes on fundamental importance. This survey covers the diverse research literature on this topic. Different categories of presentation attack are described and placed in an application-relevant framework, and the state of the art in detecting each category of attack is summarized. One conclusion from this is that presentation attack detection for iris recognition is not yet a solved problem. Datasets available for research are described, research directions for the near- and medium-term future are outlined, and a short list of recommended readings are suggested., Comment: Pre-print accepted for publication in ACM Computing Surveys on June 13, 2018

Prior art has shown it is possible to estimate, through image processing and computer vision techniques, the types and parameters of transformations that have been applied to the content of individual images to obtain new images. Given a large corpus of images and a query image, an interesting further step is to retrieve the set of original images whose content is present in the query image, as well as the detailed sequences of transformations that yield the query image given the original images. This is a problem that recently has received the name of image provenance analysis. In these times of public media manipulation ( e.g., fake news and meme sharing), obtaining the history of image transformations is relevant for fact checking and authorship verification, among many other applications. This article presents an end-to-end processing pipeline for image provenance analysis, which works at real-world scale. It employs a cutting-edge image filtering solution that is custom-tailored for the problem at hand, as well as novel techniques for obtaining the provenance graph that expresses how the images, as nodes, are ancestrally connected. A comprehensive set of experiments for each stage of the pipeline is provided, comparing the proposed solution with state-of-the-art results, employing previously published datasets. In addition, this work introduces a new dataset of real-world provenance cases from the social media site Reddit, along with baseline results., Comment: 13 pages, 6 figures

Digital retouching of face images is becoming more widespread due to the introduction of software packages that automate the task. Several researchers have introduced algorithms to detect whether a face image is original or retouched. However, previous work on this topic has not considered whether or how accuracy of retouching detection varies with the demography of face images. In this paper, we introduce a new Multi-Demographic Retouched Faces (MDRF) dataset, which contains images belonging to two genders, male and female, and three ethnicities, Indian, Chinese, and Caucasian. Further, retouched images are created using two different retouching software packages. The second major contribution of this research is a novel semi-supervised autoencoder incorporating "subclass" information to improve classification. The proposed approach outperforms existing state-of-the-art detection algorithms for the task of generalized retouching detection. Experiments conducted with multiple combinations of ethnicities show that accuracy of retouching detection can vary greatly based on the demographics of the training and testing images., Comment: Accepted in International Joint Conference on Biometrics, 2017

In this paper, we propose a novel face synthesis approach that can generate an arbitrarily large number of synthetic images of both real and synthetic identities. Thus a face image dataset can be expanded in terms of the number of identities represented and the number of images per identity using this approach, without the identity-labeling and privacy complications that come from downloading images from the web. To measure the visual fidelity and uniqueness of the synthetic face images and identities, we conducted face matching experiments with both human participants and a CNN pre-trained on a dataset of 2.6M real face images. To evaluate the stability of these synthetic faces, we trained a CNN model with an augmented dataset containing close to 200,000 synthetic faces. We used a snapshot of this trained CNN to recognize extremely challenging frontal (real) face images. Experiments showed training with the augmented faces boosted the face recognition performance of the CNN.

Iris recognition systems are a mature technology that is widely used throughout the world. In identification (as opposed to verification) mode, an iris to be recognized is typically matched against all N enrolled irises. This is the classic "1-to-N search". In order to improve the speed of large-scale identification, a modified "1-to-First" search has been used in some operational systems. A 1-to-First search terminates with the first below-threshold match that is found, whereas a 1-to-N search always finds the best match across all enrollments. We know of no previous studies that evaluate how the accuracy of 1-to-First search differs from that of 1-to-N search. Using a dataset of over 50,000 iris images from 2,800 different irises, we perform experiments to evaluate the relative accuracy of 1-to-First and 1-to-N search. We evaluate how the accuracy difference changes with larger numbers of enrolled irises, and with larger ranges of rotational difference allowed between iris images. We find that False Match error rate for 1-to-First is higher than for 1-to-N, and the the difference grows with larger number of enrolled irises and with larger range of rotation., Comment: 2016 IEEE Winter Conference on Applications of Computer Vision (WACV)

The Computer Vision Research Lab at the University of Notre Dame began collecting iris images in the spring semester of 2004. The initial data collections used an LG 2200 iris imaging system for image acquisition. Image datasets acquired in 2004-2005 at Notre Dame with this LG 2200 have been used in the ICE 2005 and ICE 2006 iris biometric evaluations. The ICE 2005 iris image dataset has been distributed to over 100 research groups around the world. The purpose of this document is to describe the content of the ND-IRIS-0405 iris image dataset. This dataset is a superset of the iris image datasets used in ICE 2005 and ICE 2006. The ND 2004-2005 iris image dataset contains 64,980 images corresponding to 356 unique subjects, and 712 unique irises. The age range of the subjects is 18 to 75 years old. 158 of the subjects are female, and 198 are male. 250 of the subjects are Caucasian, 82 are Asian, and 24 are other ethnicities., Comment: 13 pages, 8 figures