Accelerating Viola-Jones Face Detection to FPGA-Level Using GPUs
Face detection is an important aspect for biometrics, video surveillance and human computer interaction. The authors present a multi-GPU implementation of the Viola-Jones face detection algorithm that meets the performance of the fastest known FPGA implementation. The GPU design offers far lower development costs, but the FPGA implementation consumes less power. They discuss the performance programming required to realize the design, and describe future research directions. Object detection is prevalent in many applications such as security systems and bioinformatics. Whether detecting a face, mouse or stem cell, near real-time detection is essential for many applications.