Sliding-Windows for Rapid Object Class Localization: A Parallel Technique

Abstract. This paper presents a fast object class localization framework implemented on a data parallel architecture currently available in recent computers. Our case study, the implementation of Histograms of Oriented Gradients (HOG) descriptors, shows that just by using this recent programming model we can easily speed up an original CPUonly implementation by a factor of 34, making it unnecessary to use early rejection cascades that sacrice classication performance, even in real-time conditions. Using recent techniques to program the Graphics Processing Unit (GPU) allow our method to scale up to the latest, as well as to future improvements of the hardware.

• Type: Conference
• Year: 2008
• Where: DAGM
• Authors: Christian Wojek, Gyuri Dorkó, André Schulz, Bernt Schiele
• Keywords: DAGM 2008, Early Rejection Cascades, Image Processing, Object Class Localization, Original Cpuonly Implementation