The best-reviewed paper from the workshop will be invited to submit extended work to the IEEE TBIOM SPECIAL ISSUE on Best Reviewed Papers from ICCV 2025 Biometrics.
With growing global security concerns, biometric-based authentication and identification have become indispensable due to their reliability and robustness. Beyond physical biometrics, behavior understanding is emerging as a critical domain, aiming to interpret complex behavioral patterns that arise during interactions. Integrating both biometric and behavioral insights can lead to more secure, adaptive, and context-aware identity verification systems. Computer vision plays a pivotal role in analyzing and synthesizing biometric, identity, and behavior data. Recent advancements in BIB research, driven by deep learning and multimodal analysis, have significantly expanded the field. However, numerous challenges remain, including effective joint modeling of multimodal cues occurring at different time scales, handling the inherent uncertainty of machine-detectable behavioral evidence, and addressing long-term dependencies in human behavior and identity recognition. This workshop aims to bring together leading researchers, industry experts, and government agencies to discuss the latest breakthroughs in BIB research. It will serve as a platform to explore cutting-edge solutions, share innovative methodologies, and address the open challenges in this evolving field.
Call for Reviewers: We invite you to join our Technical Program Committee by participating in the paper review process for our upcoming workshop. If you are interested in contributing your expertise, please take a moment to complete the sign-up. Serving as a reviewer is an excellent opportunity to engage with cutting-edge research, collaborate with peers, and help shape the quality and direction of the workshop. All program committee members will be publicly acknowledged for their valuable contributions. Your dedication of time and knowledge is not only deeply appreciated but also essential to ensuring the success and impact of this event. We look forward to welcoming you to our team and working together to make this workshop a memorable and meaningful experience for all participants.
Reviewer's sign-up link: Google Form| Time | Session | Presenter |
|---|---|---|
| 8:00-8:05 AM | Opening Session | |
| 8:05-8:45 AM | Keynote 1 | Louis-Philippe Morency, CMU, USA |
| 8:45-9:25 AM | Keynote 2 | Stephanie Schuckers, UNCC, USA |
| 9:25-10:05 AM | Keynote 3 | Ajay Kumar, Hong Kong Polytechnic University |
| 10:05-10:20 AM | Coffee Break | |
| 10:20-11:35 AM | Oral Session 1 | Face and Generative Models |
| 11:35 AM-12:15 PM | Keynote 4 | László A. Jeni, CMU, USA |
| 12:15-1:15 PM | Oral Session 2 | Identity |
| 1:15-2:30 PM | Lunch Break | |
| 2:30-3:10 PM | Keynote 5 | Tal Hassner, WEIR AI |
| 3:10-3:25 PM | Coffee Break | |
| 3:25-4:05 PM | Keynote 6 | Amit K. Roy-Chowdhury, UC Riverside AI Research |
| 4:05-4:45 PM | Keynote 7 | Gerard Guy Medioni, Prime Video & Amazon |
| 4:45-5:00 PM | Closing Remarks |
| Session | Title | Abstract |
|---|---|---|
| Keynote 1 | Towards Embodied Social Intelligence | |
| Keynote 2 | Security and Robustness in Biometric Recognition: Challenges and Opportunities | Biometric recognition has become an everyday part of life with applications from mobile devices to air travel to finance. However, despite its relative ubiquity and success, challenges remain. Biometrics systems are vulnerable to attacks including physical artefacts such as printouts, image/video display, or masks, and, more recently, with the leap in deepfake generation tools, digital injection attacks. The ever-evolving nature of attacks makes securing against attacks challenging. Further, robustness includes ensuring biometric systems work well for all, including children. In this talk, we present our recent work on liveness, face swaps, aging in children, and fairness. |
| Keynote 3 | Unlocking the Future: Advancing AI for Trustworthy Contactless Finger & Palm Recognition | Contactless imaging of finger or palm surfaces offers unmatched user convenience, reliability, and reveals a multitude of features under various illuminations and across a broad spectrum of imaging resolutions. Pioneering spoof-resistant AI techniques capable of securely matching these patterns across least-constrained contactless variations are essential for next-gen fintech, metaverse, and emerging robotics applications. Despite their promise, significant challenges remain: ranging from the critical lack of large-scale datasets (compounded by privacy regulations like GDPR), lack of highly accurate and interpretable AI models, and the escalating threat from undetectable hyper-realistic spoofs. This talk will explore recent advancements in contactless hand-based biometric recognition and highlight the pressing challenges that must be overcome to achieve robust, scalable, and trustworthy deployment in real-world environments. |
| Keynote 4 | Towards Spatial Intelligence for Behaviors and Environments | We are in an era of foundation models and spatial intelligence (AR/VR). Despite significant advancements in natural language processing for reasoning, other modalities like vision lag behind, offering limited contributions: current video-language models (VLMs) struggle even with basic spatial and behavioral reasoning tasks. The challenge lies in the disparate training needs of different modalities. To enhance spatial reasoning, we must elevate vision to a higher semantic level (e.g., geometry), aligning it with language to achieve multimodal reasoning. Developing models that can reason about dynamic environments, behaviors, and interactions via multimodal inputs requires three key innovations: 1) universal 3D lifting to semantic representation for reasoning, 2) more efficient Vision Transformers (ViTs) for spatial tasks, and 3) robust data collection and benchmarking frameworks. In this presentation, I will discuss ongoing projects in my lab aimed at these innovations. |
| Keynote 5 | Face Recognition Under Fire: Privacy Regulations and the Great Shutdown | In the last decade, face recognition has evolved from a frontier research field into commodity technology, and recognition systems can now be deployed with just a few lines of code. Yet as the technology matured, the regulatory environment surrounding it became a minefield. Today, a system that performs flawlessly in the lab can bring a company to court, not for identifying the wrong person, but for identifying anyone without informed consent. In this talk, I trace how privacy laws like BIPA, CUBI, and GDPR reshaped the landscape, forcing Meta, Google, and others to shut down billion-user systems and pay over $2.4 billion in settlements. I will show what went wrong in those cases, what remains legally safe to build, and how the next frontier is not about improving accuracy, but about redesigning recognition systems that respect the boundaries between technical capability and legal permission. |
| Keynote 6 | Reliable Inference in Generative Models: From Data to Actions | As generative models have become more popular, some of their limitations and negative effects are also becoming apparent. On the one hand, these include malicious data generated by such models, like deepfakes. Another concern is the use of such models in safety-critical applications, and how hallucinations can affect actions that depend on data analyzed by such models. In this talk, we will consider both these concerns with generative AI. First, we will discuss how deepfakes can be detected and what the challenges are. Then, we will discuss how the uncertainty in the outputs of such models can be quantified, and how to understand hallucinations from such models when deployed with robotic agents. We will conclude by discussing future directions of research at the intersection of computer vision, machine learning, and security. |
| Keynote 7 | Prime Video: a Differentiated Viewing Experience | This talk presents an overview of the technology components powering the Prime Video customer experience. Going beyond title level information, we segment the video into shots and scenes, parse each scene to infer semantic content, and use it for a number of applications, such as content moderation, subtitles, dubbing, audio descriptions. We also augment the original content with artwork and video clips, provide cast and music recognition in X-Ray, all of which feed into the recommendation presentation. The talk ends with a presentation of AI-powered innovative features in live broadcast of sports events. |
| Time | Session | Papers & Authors |
|---|---|---|
| 10:20–11:35 AM | Oral Session 1 - Face and Generative Models |
Demographic Differentials in Face Image Quality: Evaluation and Comparison on Real and Synthetic Data Authors: André Dörsch, Johannes Merkle, Benjamin Tams, Gerardo Gutierrez Alvarez, Peter Munch, Christoph Busch, Christian Rathgeb |
|
ViT-FIQA: Assessing Face Image Quality using Vision Transformers Authors: Andrea Atzori, Fadi Boutros, Naser Damer |
||
|
On Adversarial Robustness of Face Presentation Attack Detection Algorithms Authors: Akshay Agarwal, Mayank Vatsa, Richa Singh |
||
|
TAIGen: Training-Free Adversarial Image Generation via Diffusion Models Authors: Susim Roy, Anubhooti Jain, Richa Singh, Mayank Vatsa |
||
|
Now You See Me, Now You Don’t: A Unified Framework for Expression Consistent Anonymization in Talking Head Videos Authors: Anil Egin, Andrea Tangherloni, Antitza Dantcheva |
||
| 12:15–1:15 AM | Oral Session 2 - Identity |
Intrinsically-Interpretable Siamese Networks for Identity Recognition Authors: Marco André, Jaime S Cardoso, Helena Montenegro |
|
Evaluation of Human Visual Privacy: Three-Dimensional Approach and Benchmark Dataset Authors: Sara Abdulaziz, Giacomo D'Amicantonio, Egor Bondarev |
||
|
NegFaceDiff: The Power of Negative Context in Identity-Conditioned Diffusion for Synthetic Face Generation Authors: Eduarda Caldeira, Naser Damer, Fadi Boutros |
||
|
Are you In or Out (of gallery)? Wisdom from the Same-Identity Crowd Authors: Aman Bhatta, Maria Dhakal, Michael C. King, Kevin Bowyer |
||