Isotropic Granularity-tunable gradients partition (IGGP) descriptors for human detection

• DOI: 10.5244/c.24.63
• Publication Date: 2010-08-31T07:08:29Z
• Authors: Yazhou Liu; Janne Heikkila
• Keywords: Granularity; Speedup; Benchmark (surveying)

This paper presents a new descriptor for human detection in still images. It is referred to as isotropic granularity-tunable gradients partition (IGGP), which extended from (GGP) descriptors. The representation achieved by aligning the features with different orientation channels according their principal angles. benefits of this extension are two folds: firstly, since partitions’ sizes all equal, noise introduce small partitions original GGP descriptors eliminated and performance can be essentially improved; secondly, integral image based fast computation applied more than 20 times speedup has been achieved. In addition, we dataset HIMA. Unlike previous available datasets mainly captured on street views automobile safety or robotics, HIMA outdoor work fields industry safety. major challenges include: extreme light conditions, occlusion strong noise. We benchmark several promising systems, providing an overview state-of-the-art set. Experimental results show that proposed method yield very competitive both speed accuracy.