1. What are the contributions mentioned in the paper "Fast triangular binning kernel approximations for weighted gradient histogram creation" ?
The implementation of weighted gradient histograms are studied.. The authors show that previously presented fast methods for uniform binning kernels can be extended to non-uniform binning, and that the triangular kernel can be well approximated for common weighting strategies.. The approximation is implemented with sums and products of projections of the gradient samples on specially chosen vectors.. Finally, the frequency components of the different kernels are studied to quantify the fundamental gain achieved by using triangular kernels instead of uniform kernels.. Further, due to potentially severe aliasing, non-uniform binning kernels are desirable.. Consequently, only a few standard arithmetic operations are required, and therefore, the suggested implementation has a significantly lower computational cost when compared with an implementation in which the gradient argument and magnitude are explicitly evaluated.
read more
2. What are the future works in "Fast triangular binning kernel approximations for weighted gradient histogram creation" ?
At first this may seem to eliminate the possibility to use existing fast implementations of the binning ; however, in this article the authors have extended earlier presented fast implementation methods and shown that the triangular kernel can be well approximated together with common magnitude weighting strategies.
read more
![Fig. 3: Weighting functions (2) and (3) and (4) and (5) for 8 bins plotted over a range extending outside their applicable regions. Outside the region marked with the dashed lines, the bins j and j+1 will not be the closest pair of bins. The binning methods in [4] are seen to select approximately linear portions of the functions.](/figures/fig-3-weighting-functions-2-and-3-and-4-and-5-for-8-bins-1ndw33lj.webp)