12th IEEE International Conference on Advanced Video and Signal based Surveillance Karlsruhe, Germany — August 25–28, 2015
A clean architectural photograph with strong symmetry: a modern university lecture hall interior with tiered seating, warm wood accents, and soft overhead lighting

Keynote Lectures

Insights from leading researchers in computer vision and signal-based surveillance.

The IEEE AVSS conference featured keynote presentations from distinguished researchers at the forefront of video and signal-based surveillance. These lectures addressed foundational challenges and emerging directions in the field.

The keynote lectures at this conference bring together leading researchers from around the globe to share groundbreaking advances in computer vision and signal-based surveillance. Each presentation offers attendees a deep dive into the latest theoretical frameworks and practical methodologies shaping the field. From intelligent video analytics to robust activity recognition, the talks cover a wide spectrum of topics that address both current challenges and future opportunities. These sessions are designed to spark meaningful discussions among academia, industry, and government representatives, fostering collaboration that drives the next generation of surveillance technologies forward.

A central theme across many keynote addresses is the integration of advanced machine learning techniques with real-world monitoring systems. Speakers explore how deep learning models are being refined to improve accuracy in object detection, tracking, and behavior analysis under complex conditions. Emphasis is placed on developing algorithms that can operate reliably across diverse environments, from crowded urban settings to remote border areas. Researchers also highlight the importance of balancing performance with ethical considerations, ensuring that technological progress in surveillance goes hand in hand with responsible deployment and respect for privacy.

The conference program includes keynote lectures that examine the evolution of video-based surveillance beyond traditional security applications. Presenters discuss how these technologies are being adapted for traffic management, vehicle tracking, and industrial process control, demonstrating their versatility in solving pressing societal problems. Attendees gain insight into how multimodal data sources, including signal processing and biometric systems, are being fused to create more comprehensive monitoring solutions. These interdisciplinary approaches open new avenues for research and practical implementation, making the keynote sessions a vital component of the overall conference experience.

Beyond the technical content, the keynote lectures serve as a platform for visionary thinking about the future of surveillance science. Speakers challenge audiences to consider how emerging trends such as edge computing, real-time analytics, and autonomous decision-making will reshape the landscape. The discussions encourage participants to think critically about scalability, robustness, and the role of human oversight in automated systems. By presenting both successful case studies and lessons learned from failures, the keynotes provide a balanced perspective that helps researchers and practitioners navigate the complexities of this rapidly advancing domain.

How and Why to Learn a 3D Model of the Human Body in Motion

Michael J. Black — MPI for Intelligent Systems, Tübingen, Germany

Abstract

The human body is complex and deformable. For many applications in computer vision, graphics, fashion, and medicine, having a realistic, low-dimensional, 3D model of the body is useful. Getting such a model, however, is challenging. This keynote explored methods for learning 3D body models from data and their significance for advancing surveillance and related domains.