Event cameras, also known as Dynamic Vision Sensors (DVS), are an emerging visual modality
that is attracting significant interest in the fields of robotics and computer vision. These cameras
are distinguished by their exceptionally high temporal resolution and dynamic range, low yet
adaptive power consumption, sparse output, and a dynamic vision scheme akin to mammalian
perception. Such attributes have enabled their success in various computer vision applications,
including feature tracking, optical flow estimation, and pose estimation.
As a promising and rapidly growing field, the body of research on event cameras has expanded
significantly in recent years, and their applications are becoming increasingly widespread. The
spike-like characteristics of event cameras make them an ideal visual modality for spiking neural
networks and graph neural networks, enhancing their potential for next-generation computing.
In this talk, I will provide an overview of the domain of event cameras, highlighting their unique
characteristics, diverse applications, and potential future developments. Additionally, I will
present some of our featured works in this field.