Kai Cao

Kanishka P. Wijewardena, Steven A. Grosz, Kai Cao et al.

Much of the success of fingerprint recognition is attributed to minutiae-based fingerprint representation. It was believed that minutiae templates could not be inverted to obtain a high fidelity fingerprint image, but this assumption has been shown to be false. The success of deep learning has resulted in alternative fingerprint representations (embeddings), in the hope that they might offer better recognition accuracy as well as non-invertibility of deep network-based templates. We evaluate whether deep fingerprint templates suffer from the same reconstruction attacks as the minutiae templates. We show that while a deep template can be inverted to produce a fingerprint image that could be matched to its source image, deep templates are more resistant to reconstruction attacks than minutiae templates. In particular, reconstructed fingerprint images from minutiae templates yield a TAR of about 100.0% (98.3%) @ FAR of 0.01% for type-I (type-II) attacks using a state-of-the-art commercial fingerprint matcher, when tested on NIST SD4. The corresponding attack performance for reconstructed fingerprint images from deep templates using the same commercial matcher yields a TAR of less than 1% for both type-I and type-II attacks; however, when the reconstructed images are matched using the same deep network, they achieve a TAR of 85.95% (68.10%) for type-I (type-II) attacks. Furthermore, what is missing from previous fingerprint template inversion studies is an evaluation of the black-box attack performance, which we perform using 3 different state-of-the-art fingerprint matchers. We conclude that fingerprint images generated by inverting minutiae templates are highly susceptible to both white-box and black-box attack evaluations, while fingerprint images generated by deep templates are resistant to black-box evaluations and comparatively less susceptible to white-box evaluations.

Joshua J. Engelsma, Kai Cao, Anil K. Jain

We open source fingerprint Match in Box, a complete end-to-end fingerprint recognition system embedded within a 4 inch cube. Match in Box stands in contrast to a typical bulky and expensive proprietary fingerprint recognition system which requires sending a fingerprint image to an external host for processing and subsequent spoof detection and matching. In particular, Match in Box is a first of a kind, portable, low-cost, and easy-to-assemble fingerprint reader with an enrollment database embedded within the reader's memory and open source fingerprint spoof detector, feature extractor, and matcher all running on the reader's internal vision processing unit (VPU). An onboard touch screen and rechargeable battery pack make this device extremely portable and ideal for applying both fingerprint authentication (1:1 comparison) and fingerprint identification (1:N search) to applications (vaccination tracking, food and benefit distribution programs, human trafficking prevention) in rural communities, especially in developing countries. We also show that Match in Box is suited for capturing neonate fingerprints due to its high resolution (1900 ppi) cameras.

Dinh-Luan Nguyen, Kai Cao, Anil K. Jain

We propose a fully automatic minutiae extractor, called MinutiaeNet, based on deep neural networks with compact feature representation for fast comparison of minutiae sets. Specifically, first a network, called CoarseNet, estimates the minutiae score map and minutiae orientation based on convolutional neural network and fingerprint domain knowledge (enhanced image, orientation field, and segmentation map). Subsequently, another network, called FineNet, refines the candidate minutiae locations based on score map. We demonstrate the effectiveness of using the fingerprint domain knowledge together with the deep networks. Experimental results on both latent (NIST SD27) and plain (FVC 2004) public domain fingerprint datasets provide comprehensive empirical support for the merits of our method. Further, our method finds minutiae sets that are better in terms of precision and recall in comparison with state-of-the-art on these two datasets. Given the lack of annotated fingerprint datasets with minutiae ground truth, the proposed approach to robust minutiae detection will be useful to train network-based fingerprint matching algorithms as well as for evaluating fingerprint individuality at scale. MinutiaeNet is implemented in Tensorflow: https://github.com/luannd/MinutiaeNet

Joshua J. Engelsma, Kai Cao, Anil K. Jain

We open source an easy to assemble, spoof resistant, high resolution, optical fingerprint reader, called RaspiReader, using ubiquitous components. By using our open source STL files and software, RaspiReader can be built in under one hour for only US $175. As such, RaspiReader provides the fingerprint research community a seamless and simple method for quickly prototyping new ideas involving fingerprint reader hardware. In particular, we posit that this open source fingerprint reader will facilitate the exploration of novel fingerprint spoof detection techniques involving both hardware and software. We demonstrate one such spoof detection technique by specially customizing RaspiReader with two cameras for fingerprint image acquisition. One camera provides high contrast, frustrated total internal reflection (FTIR) fingerprint images, and the other outputs direct images of the finger in contact with the platen. Using both of these image streams, we extract complementary information which, when fused together and used for spoof detection, results in marked performance improvement over previous methods relying only on grayscale FTIR images provided by COTS optical readers. Finally, fingerprint matching experiments between images acquired from the FTIR output of RaspiReader and images acquired from a COTS reader verify the interoperability of the RaspiReader with existing COTS optical readers.

Joshua J. Engelsma, Kai Cao, Anil K. Jain

We present the design and prototype of an open source, optical fingerprint reader, called RaspiReader, using ubiquitous components. RaspiReader, a low-cost and easy to assemble reader, provides the fingerprint research community a seamless and simple method for gaining more control over the sensing component of fingerprint recognition systems. In particular, we posit that this versatile fingerprint reader will encourage researchers to explore novel spoof detection methods that integrate both hardware and software. RaspiReader's hardware is customized with two cameras for fingerprint acquisition with one camera providing high contrast, frustrated total internal reflection (FTIR) images, and the other camera outputting direct images. Using both of these image streams, we extract complementary information which, when fused together, results in highly discriminative features for fingerprint spoof (presentation attack) detection. Our experimental results demonstrate a marked improvement over previous spoof detection methods which rely only on FTIR images provided by COTS optical readers. Finally, fingerprint matching experiments between images acquired from the FTIR output of the RaspiReader and images acquired from a COTS fingerprint reader verify the interoperability of the RaspiReader with existing COTS optical readers.

Wenqi Zhou, Kai Cao, Hao Zheng et al.

Long-form egocentric video understanding provides rich contextual information and unique insights into long-term human behaviors, holding significant potential for applications in embodied intelligence, long-term activity analysis, and personalized assistive technologies. However, existing benchmark datasets primarily focus on single, short (\eg, minutes to tens of minutes) to moderately long videos, leaving a substantial gap in evaluating extensive, ultra-long egocentric video recordings. To address this, we introduce X-LeBench, a novel benchmark dataset meticulously designed to fill this gap by focusing on tasks requiring a comprehensive understanding of extremely long egocentric video recordings. Our X-LeBench develops a life-logging simulation pipeline that produces realistic, coherent daily plans aligned with real-world video data. This approach enables the flexible integration of synthetic daily plans with real-world footage from Ego4D-a massive-scale egocentric video dataset covers a wide range of daily life scenarios-resulting in 432 simulated video life logs spanning from 23 minutes to 16.4 hours. The evaluations of several baseline systems and multimodal large language models (MLLMs) reveal their poor performance across the board, highlighting the inherent challenges of long-form egocentric video understanding, such as temporal localization and reasoning, context aggregation, and memory retention, and underscoring the need for more advanced models.

Tianyi Zhao, Kai Cao, Jiawen Yao et al.

The pancreatic disease taxonomy includes ten types of masses (tumors or cysts)[20,8]. Previous work focuses on developing segmentation or classification methods only for certain mass types. Differential diagnosis of all mass types is clinically highly desirable [20] but has not been investigated using an automated image understanding approach. We exploit the feasibility to distinguish pancreatic ductal adenocarcinoma (PDAC) from the nine other nonPDAC masses using multi-phase CT imaging. Both image appearance and the 3D organ-mass geometry relationship are critical. We propose a holistic segmentation-mesh-classification network (SMCN) to provide patient-level diagnosis, by fully utilizing the geometry and location information, which is accomplished by combining the anatomical structure and the semantic detection-by-segmentation network. SMCN learns the pancreas and mass segmentation task and builds an anatomical correspondence-aware organ mesh model by progressively deforming a pancreas prototype on the raw segmentation mask (i.e., mask-to-mesh). A new graph-based residual convolutional network (Graph-ResNet), whose nodes fuse the information of the mesh model and feature vectors extracted from the segmentation network, is developed to produce the patient-level differential classification results. Extensive experiments on 661 patients' CT scans (five phases per patient) show that SMCN can improve the mass segmentation and detection accuracy compared to the strong baseline method nnUNet (e.g., for nonPDAC, Dice: 0.611 vs. 0.478; detection rate: 89% vs. 70%), achieve similar sensitivity and specificity in differentiating PDAC and nonPDAC as expert radiologists (i.e., 94% and 90%), and obtain results comparable to a multimodality test [20] that combines clinical, imaging, and molecular testing for clinical management of patients.

Joshua J. Engelsma, Debayan Deb, Kai Cao et al.

In many of the least developed and developing countries, a multitude of infants continue to suffer and die from vaccine-preventable diseases and malnutrition. Lamentably, the lack of official identification documentation makes it exceedingly difficult to track which infants have been vaccinated and which infants have received nutritional supplements. Answering these questions could prevent this infant suffering and premature death around the world. To that end, we propose Infant-Prints, an end-to-end, low-cost, infant fingerprint recognition system. Infant-Prints is comprised of our (i) custom built, compact, low-cost (85 USD), high-resolution (1,900 ppi), ergonomic fingerprint reader, and (ii) high-resolution infant fingerprint matcher. To evaluate the efficacy of Infant-Prints, we collected a longitudinal infant fingerprint database captured in 4 different sessions over a 12-month time span (December 2018 to January 2020), from 315 infants at the Saran Ashram Hospital, a charitable hospital in Dayalbagh, Agra, India. Our experimental results demonstrate, for the first time, that Infant-Prints can deliver accurate and reliable recognition (over time) of infants enrolled between the ages of 2-3 months, in time for effective delivery of vaccinations, healthcare, and nutritional supplements (TAR=95.2% @ FAR = 1.0% for infants aged 8-16 weeks at enrollment and authenticated 3 months later).

Ling Zhang, Yu Shi, Jiawen Yao et al.

Accurate and automated tumor segmentation is highly desired since it has the great potential to increase the efficiency and reproducibility of computing more complete tumor measurements and imaging biomarkers, comparing to (often partial) human measurements. This is probably the only viable means to enable the large-scale clinical oncology patient studies that utilize medical imaging. Deep learning approaches have shown robust segmentation performances for certain types of tumors, e.g., brain tumors in MRI imaging, when a training dataset with plenty of pixel-level fully-annotated tumor images is available. However, more than often, we are facing the challenge that only (very) limited annotations are feasible to acquire, especially for hard tumors. Pancreatic ductal adenocarcinoma (PDAC) segmentation is one of the most challenging tumor segmentation tasks, yet critically important for clinical needs. Previous work on PDAC segmentation is limited to the moderate amounts of annotated patient images (n<300) from venous or venous+arterial phase CT scans. Based on a new self-learning framework, we propose to train the PDAC segmentation model using a much larger quantity of patients (n~=1,000), with a mix of annotated and un-annotated venous or multi-phase CT images. Pseudo annotations are generated by combining two teacher models with different PDAC segmentation specialties on unannotated images, and can be further refined by a teaching assistant model that identifies associated vessels around the pancreas. A student model is trained on both manual and pseudo annotated multi-phase images. Experiment results show that our proposed method provides an absolute improvement of 6.3% Dice score over the strong baseline of nnUNet trained on annotated images, achieving the performance (Dice = 0.71) similar to the inter-observer variability between radiologists.

Joshua J. Engelsma, Kai Cao, Anil K. Jain

We present DeepPrint, a deep network, which learns to extract fixed-length fingerprint representations of only 200 bytes. DeepPrint incorporates fingerprint domain knowledge, including alignment and minutiae detection, into the deep network architecture to maximize the discriminative power of its representation. The compact, DeepPrint representation has several advantages over the prevailing variable length minutiae representation which (i) requires computationally expensive graph matching techniques, (ii) is difficult to secure using strong encryption schemes (e.g. homomorphic encryption), and (iii) has low discriminative power in poor quality fingerprints where minutiae extraction is unreliable. We benchmark DeepPrint against two top performing COTS SDKs (Verifinger and Innovatrics) from the NIST and FVC evaluations. Coupled with a re-ranking scheme, the DeepPrint rank-1 search accuracy on the NIST SD4 dataset against a gallery of 1.1 million fingerprints is comparable to the top COTS matcher, but it is significantly faster (DeepPrint: 98.80% in 0.3 seconds vs. COTS A: 98.85% in 27 seconds). To the best of our knowledge, the DeepPrint representation is the most compact and discriminative fixed-length fingerprint representation reported in the academic literature.

Joshua J. Engelsma, Kai Cao, Anil K. Jain

We learn a discriminative fixed length feature representation of fingerprints which stands in contrast to commonly used unordered, variable length sets of minutiae points. To arrive at this fixed length representation, we embed fingerprint domain knowledge into a multitask deep convolutional neural network architecture. Empirical results, on two public-domain fingerprint databases (NIST SD4 and FVC 2004 DB1) show that compared to minutiae representations, extracted by two state-of-the-art commercial matchers (Verifinger v6.3 and Innovatrics v2.0.3), our fixed-length representations provide (i) higher search accuracy: Rank-1 accuracy of 97.9% vs. 97.3% on NIST SD4 against a gallery size of 2000 and (ii) significantly faster, large scale search: 682,594 matches per second vs. 22 matches per second for commercial matchers on an i5 3.3 GHz processor with 8 GB of RAM.

Kai Cao, Dinh-Luan Nguyen, Cori Tymoszek et al.

Latent fingerprints are one of the most important and widely used sources of evidence in law enforcement and forensic agencies. Yet the performance of the state-of-the-art latent recognition systems is far from satisfactory, and they often require manual markups to boost the latent search performance. Further, the COTS systems are proprietary and do not output the true comparison scores between a latent and reference prints to conduct quantitative evidential analysis. We present an end-to-end latent fingerprint search system, including automated region of interest (ROI) cropping, latent image preprocessing, feature extraction, feature comparison , and outputs a candidate list. Two separate minutiae extraction models provide complementary minutiae templates. To compensate for the small number of minutiae in small area and poor quality latents, a virtual minutiae set is generated to construct a texture template. A 96-dimensional descriptor is extracted for each minutia from its neighborhood. For computational efficiency, the descriptor length for virtual minutiae is further reduced to 16 using product quantization. Our end-to-end system is evaluated on three latent databases: NIST SD27 (258 latents); MSP (1,200 latents), WVU (449 latents) and N2N (10,000 latents) against a background set of 100K rolled prints, which includes the true rolled mates of the latents with rank-1 retrieval rates of 65.7%, 69.4%, 65.5%, and 7.6% respectively. A multi-core solution implemented on 24 cores obtains 1ms per latent to rolled comparison.

Kai Cao, Anil K. Jain

We propose a texture template approach, consisting of a set of virtual minutiae, to improve the overall latent fingerprint recognition accuracy. To compensate for the lack of sufficient number of minutiae in poor quality latent prints, we generate a set of virtual minutiae. However, due to a large number of these regularly placed virtual minutiae, texture based template matching has a large computational requirement compared to matching true minutiae templates. To improve both the accuracy and efficiency of the texture template matching, we investigate: i) both original and enhanced fingerprint patches for training convolutional neural networks (ConvNets) to improve the distinctiveness of descriptors associated with each virtual minutiae, ii) smaller patches around virtual minutiae and a fast ConvNet architecture to speed up descriptor extraction, iii) reduce the descriptor length, iv) a modified hierarchical graph matching strategy to improve the matching speed, and v) extraction of multiple texture templates to boost the performance. Experiments on NIST SD27 latent database show that the above strategies can improve the matching speed from 11 ms (24 threads) per comparison (between a latent and a reference print) to only 7.7 ms (single thread) per comparison while improving the rank-1 accuracy by 8.9% against 10K gallery.

Dinh-Luan Nguyen, Kai Cao, Anil K. Jain

We present a simple but effective method for automatic latent fingerprint segmentation, called SegFinNet. SegFinNet takes a latent image as an input and outputs a binary mask highlighting the friction ridge pattern. Our algorithm combines fully convolutional neural network and detection-based approaches to process the entire input latent image in one shot instead of using latent patches. Experimental results on three different latent databases (i.e. NIST SD27, WVU, and an operational forensic database) show that SegFinNet outperforms both human markup for latents and the state-of-the-art latent segmentation algorithms. We further show that this improved cropping boosts the hit rate of a latent fingerprint matcher.

Debayan Deb, Tarang Chugh, Joshua Engelsma et al.

We address the problem of comparing fingerphotos, fingerprint images from a commodity smartphone camera, with the corresponding legacy slap contact-based fingerprint images. Development of robust versions of these technologies would enable the use of the billions of standard Android phones as biometric readers through a simple software download, dramatically lowering the cost and complexity of deployment relative to using a separate fingerprint reader. Two fingerphoto apps running on Android phones and an optical slap reader were utilized for fingerprint collection of 309 subjects who primarily work as construction workers, farmers, and domestic helpers. Experimental results show that a True Accept Rate (TAR) of 95.79 at a False Accept Rate (FAR) of 0.1% can be achieved in matching fingerphotos to slaps (two thumbs and two index fingers) using a COTS fingerprint matcher. By comparison, a baseline TAR of 98.55% at 0.1% FAR is achieved when matching fingerprint images from two different contact-based optical readers. We also report the usability of the two smartphone apps, in terms of failure to acquire rate and fingerprint acquisition time. Our results show that fingerphotos are promising to authenticate individuals (against a national ID database) for banking, welfare distribution, and healthcare applications in developing countries.

Tarang Chugh, Kai Cao, Anil K. Jain

The primary purpose of a fingerprint recognition system is to ensure a reliable and accurate user authentication, but the security of the recognition system itself can be jeopardized by spoof attacks. This study addresses the problem of developing accurate, generalizable, and efficient algorithms for detecting fingerprint spoof attacks. Specifically, we propose a deep convolutional neural network based approach utilizing local patches centered and aligned using fingerprint minutiae. Experimental results on three public-domain LivDet datasets (2011, 2013, and 2015) show that the proposed approach provides state-of-the-art accuracies in fingerprint spoof detection for intra-sensor, cross-material, cross-sensor, as well as cross-dataset testing scenarios. For example, in LivDet 2015, the proposed approach achieves 99.03% average accuracy over all sensors compared to 95.51% achieved by the LivDet 2015 competition winners. Additionally, two new fingerprint presentation attack datasets containing more than 20,000 images, using two different fingerprint readers, and over 12 different spoof fabrication materials are collected. We also present a graphical user interface, called Fingerprint Spoof Buster, that allows the operator to visually examine the local regions of the fingerprint highlighted as live or spoof, instead of relying on only a single score as output by the traditional approaches.

Kai Cao, Anil K. Jain

Latent fingerprints are one of the most important and widely used evidence in law enforcement and forensic agencies worldwide. Yet, NIST evaluations show that the performance of state-of-the-art latent recognition systems is far from satisfactory. An automated latent fingerprint recognition system with high accuracy is essential to compare latents found at crime scenes to a large collection of reference prints to generate a candidate list of possible mates. In this paper, we propose an automated latent fingerprint recognition algorithm that utilizes Convolutional Neural Networks (ConvNets) for ridge flow estimation and minutiae descriptor extraction, and extract complementary templates (two minutiae templates and one texture template) to represent the latent. The comparison scores between the latent and a reference print based on the three templates are fused to retrieve a short candidate list from the reference database. Experimental results show that the rank-1 identification accuracies (query latent is matched with its true mate in the reference database) are 64.7% for the NIST SD27 and 75.3% for the WVU latent databases, against a reference database of 100K rolled prints. These results are the best among published papers on latent recognition and competitive with the performance (66.7% and 70.8% rank-1 accuracies on NIST SD27 and WVU DB, respectively) of a leading COTS latent Automated Fingerprint Identification System (AFIS). By score-level (rank-level) fusion of our system with the commercial off-the-shelf (COTS) latent AFIS, the overall rank-1 identification performance can be improved from 64.7% and 75.3% to 73.3% (74.4%) and 76.6% (78.4%) on NIST SD27 and WVU latent databases, respectively.

Guangcan Mai, Kai Cao, Pong C. Yuen et al.

State-of-the-art face recognition systems are based on deep (convolutional) neural networks. Therefore, it is imperative to determine to what extent face templates derived from deep networks can be inverted to obtain the original face image. In this paper, we study the vulnerabilities of a state-of-the-art face recognition system based on template reconstruction attack. We propose a neighborly de-convolutional neural network (\textit{NbNet}) to reconstruct face images from their deep templates. In our experiments, we assumed that no knowledge about the target subject and the deep network are available. To train the \textit{NbNet} reconstruction models, we augmented two benchmark face datasets (VGG-Face and Multi-PIE) with a large collection of images synthesized using a face generator. The proposed reconstruction was evaluated using type-I (comparing the reconstructed images against the original face images used to generate the deep template) and type-II (comparing the reconstructed images against a different face image of the same subject) attacks. Given the images reconstructed from \textit{NbNets}, we show that for verification, we achieve TAR of 95.20\% (58.05\%) on LFW under type-I (type-II) attacks @ FAR of 0.1\%. Besides, 96.58\% (92.84\%) of the images reconstruction from templates of partition \textit{fa} (\textit{fb}) can be identified from partition \textit{fa} in color FERET. Our study demonstrates the need to secure deep templates in face recognition systems.

Anil K. Jain, Sunpreet S. Arora, Lacey Best-Rowden et al.

With a number of emerging applications requiring biometric recognition of children (e.g., tracking child vaccination schedules, identifying missing children and preventing newborn baby swaps in hospitals), investigating the temporal stability of biometric recognition accuracy for children is important. The persistence of recognition accuracy of three of the most commonly used biometric traits (fingerprints, face and iris) has been investigated for adults. However, persistence of biometric recognition accuracy has not been studied systematically for children in the age group of 0-4 years. Given that very young children are often uncooperative and do not comprehend or follow instructions, in our opinion, among all biometric modalities, fingerprints are the most viable for recognizing children. This is primarily because it is easier to capture fingerprints of young children compared to other biometric traits, e.g., iris, where a child needs to stare directly towards the camera to initiate iris capture. In this report, we detail our initiative to investigate the persistence of fingerprint recognition for children in the age group of 0-4 years. Based on preliminary results obtained for the data collected in the first phase of our study, use of fingerprints for recognition of 0-4 year-old children appears promising.