Kernels on Structured Objects Through Nested Histograms
نویسندگان
چکیده
We propose a family of kernels for structured objectswhich is based on the bag-of-components paradigm. Rather than decomposing each complex object into the single histogram of its components, we use for each object a family of nested histograms, where each histogram in this hierarchy describes the object seen from an increasingly granular perspective. We use this hierarchy of histograms to define elementary kernels which can detect coarse and fine similarities between the objects. We compute through an efficient averaging trick a mixture of such specific kernels, to propose a final kernel value which weights efficiently local and global matches. We propose experimental results on an image retrieval experiment which show that that our approach is effective and can be seen as a template procedure to be used with kernels on histograms.
منابع مشابه
Multiresolution Kernels
We present in this work a new methodology to design kernels on data which is structured with smaller components, such as text, images or sequences. This methodology is a template procedure which can be applied on most kernels on measures and takes advantage of a more detailed “bag of components” representation of the objects. To obtain such a detailed description, we consider possible decomposi...
متن کامل0 50 70 33 v 2 1 4 N ov 2 00 5 Multiresolution Kernels
We present in this work a new methodology to design kernels on data which is structured with smaller components, such as text, images or sequences. This methodology is a template procedure which can be applied on most kernels on measures and takes advantage of a more detailed “bag of components” representation of the objects. To obtain such a detailed description, we consider possible decomposi...
متن کاملar X iv : c s / 05 07 03 3 v 1 [ cs . L G ] 1 3 Ju l 2 00 5 Multiresolution Kernels
We present in this work a new methodology to design kernels on data which is structured with smaller components, such as text, images or sequences. This methodology is a template procedure which can be applied on most kernels on measures and takes advantage of a more detailed “bag of components” representation of the objects. To obtain such a detailed description, we consider possible decomposi...
متن کاملCharacteristic Kernels on Structured Domains Excel in Robotics and Human Action Recognition
Embedding probability distributions into a sufficiently rich (characteristic) reproducing kernel Hilbert space enables us to take higher order statistics into account. Characterization also retains effective statistical relation between inputs and outputs in regression and classification. Recent works established conditions for characteristic kernels on groups and semigroups. Here we study char...
متن کاملHarmonic Shape Histograms for 3D Shape Classification and Retrieval
In this paper, we present a novel approach towards 3D shape recognition and retrieval using histograms of rotation invariant local features. Features are extracted for every point of voxelized 3D shape objects by use of functions on spheres which are invariant towards rotation of the object. The fast computation of the local features is performed via convolution methods in frequency space. Hist...
متن کامل