Competition#1: Latent in the Wild Fingerprint Recognition Competition
In the past decade, several studies have focused on developing latent fingerprint recognition algorithms.However, the performance evaluation of these methods was conducted on only a few databases. Despite being valuable, these databases have major shortcomings: 1) a small number of subjects respectively finger instances and latent fingerprint samples, 2) a constrained acquisition environment, and 3) limited availability. Moreover, one of the most commonly used latent fingerprint databases NIST SD27 Database has been withdrawn, making the development and performance evaluation of latent fingerprint recognition even more difficult. From a review of existing works, we note that latent fingerprint recognition algorithms have rarely been tested on large-scale datasets. To the best of our knowledge, there is no large-scale latent fingerprint in the wild database containing reference fingerprints (ground truth), latent fingerprints, and fingerphotos acquired from different surfaces that come from a large number of unique subjects. Therefore, it is necessary to establish a latent fingerprint comparison competition that evaluate comparison algorithms on a new large-scale latent fingerprint in the wild database to meet the need for robust latent fingerprint recognition algorithm development and evaluation.
Website: https://sites.google.com/view/ijcb-latentinthewild-2024
Competition#2: LivDet-Face 2024 Competition
LivDet-Face 2024, in its second iteration, stands as a cornerstone in face-liveness competitions, spotlighting the pivotal role of Presentation Attack Detection (PAD) in the overarching reliability of face recognition systems. The competition holds a dual significance, serving as both a comprehensive assessment of the current state-of-the-art in face PAD and an invaluable resource for the research community. Conducted biennially, LivDet-Face 2024 carries a strategic objective: to evaluate and push the boundaries of face presentation attack detection technologies. It accomplishes this by offering an established evaluation protocol with datasets featuring a diverse range of spoof and live face images. Beyond the competition’s timeframe, this protocol endures as a valuable reference for researchers, providing a standardized framework for assessing and comparing their solutions against the benchmarks set by LivDet-Face winners. By fostering this recurrent event, LivDet-Face aspires to perpetuate advancements in face presentation attack detection, elevating the robustness and reliability of face recognition systems in the ever-evolving landscape of biometric security.
Website: https://face2024.livdet.org//
Competition#3: UCCS Watchlist Challenge: 3rd Open-set Face Detection and Identification
Join the UCCS Watchlist Challenge, an unprecedented opportunity to revolutionize face detection and recognition for real-world surveillance applications! Are you ready to push the boundaries of biometric technology? In this cutting-edge competition, participants will navigate the complexities of open-set face detection and identification using the unique UnConstrained College Students (UCCS) dataset. This upgraded dataset offers a unique chance to investigate the capabilities of biometric systems in detecting and identifying faces under the most challenging conditions, including significant blur, occlusions, varied facial orientations, and different weather conditions. Moreover, our challenge addresses the critical need for robust solutions in situations where it is crucial to identify individuals (persons of interest) on the watchlist while rejecting others (innocents) – mirroring real-world surveillance scenarios. The challenge is split into two engaging parts: detecting faces without concern for identity, and then the thrilling task of matching these faces against subjects of a watchlist. The first requires participants to detect every face captured within the UCCS images, setting the stage for the second part. Here, the task evolves into accurately spotting watchlist individuals while rejecting those who are categorized as “unknown”. This pivotal task underscores the challenge’s core goal: to enhance the precision and reliability of open-set recognition systems, where the ability to distinguish between watchlist individuals and “unknowns” is emphasized. Therefore, this challenge is not just a competition; it’s a call to innovate in the area of surveillance-based face recognition. Whether you’re an experienced researcher or a newcomer in the field of biometrics, we invite academia’s brightest minds to dive into this competition, which aims not only to benchmark the robustness and performance of facial recognition algorithms but also to foster a collaborative environment. Don’t miss out on the chance to showcase your skills, collaborate with peers, and help shape the future of facial recognition on a global stage! Join us at IJCB 2024 to be part of an exciting journey. The challenge awaits; are you up for it?
Website: https://www.ifi.uzh.ch/en/aiml/challenge.html
Competition#4: First Competition on Presentation Attack Detection on ID-Card
The accelerated evolution in consumer smartphone cameras has increased the industry’s interest in remote biometric verification systems. The capacity to reach the customer remotely for services such as e-commerce, digital banking, and general fintech requires robust systems for remote identity verification. One approach for this verification is using an official identity document, such as a national ID card, and comparing the data with a frontal face photograph (selfie) of the person in question, also provided remotely by the user. Today, the raising of attacks on this PAD ID card system is increasing every day, and the number of images available for training and testing is limited because of privacy concerns. As a result, many solutions are over-fitted on intra-dataset therefore, limiting the generalization capabilities. PAD-ID Card 2024 will be the first competition in the ID Card series, offering: (a) an independent assessment of current state-of-the-art on ID Card Presentation Attack Detection algorithms and (b) an evaluation protocol, including a real dataset of attacks and bona fide ID card images, that can be followed by researchers after the competition is closed to benchmark their solutions with ID cards winners and baselines. Today, there is no available independent evaluation on cross-dataset with real data that allows us to know the state-of-the-art status.
Website: https://sites.google.com/view/ijcb-pad-id-card-2024/home
Competition#5: The 5th International Competition on Human Identification at a Distance 2024 (HID 2024)
The competition focuses on human identification at a distance (HID) in videos. The competition was designed to promote the research on HID. All participants from academic and industry are welcome. CodaLab is used for submissions and evaluations. The dataset for evaluation is SUSTech-Competition, a new dataset collected in 2022 and has been used in HID 2023. It contains 859 subjects. The sponsor, Watrix Technology, will provide 6 awards (19,000 CNY in total, ~2,660 USD) to the top 6 teams in the second phase. We are grateful to Watrix Technology to sponsor the competition.
Website: https://hid2024.iapr-tc4.org