Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification
In this paper we present some extensions to the k-means algorithm for vector quantization that permit its efficient use in image segmentation and pattern classification tasks. It is shown that by introducing state variables that correspond to certain statistics of the dynamic behavior of the a...
| Main Authors: | , |
|---|---|
| Language: | en_US |
| Published: |
2004
|
| Online Access: | http://hdl.handle.net/1721.1/6613 |
| _version_ | 1811091971328966656 |
|---|---|
| author | Marroquin, Jose L. Girosi, Federico |
| author_facet | Marroquin, Jose L. Girosi, Federico |
| author_sort | Marroquin, Jose L. |
| collection | MIT |
| description | In this paper we present some extensions to the k-means algorithm for vector quantization that permit its efficient use in image segmentation and pattern classification tasks. It is shown that by introducing state variables that correspond to certain statistics of the dynamic behavior of the algorithm, it is possible to find the representative centers fo the lower dimensional maniforlds that define the boundaries between classes, for clouds of multi-dimensional, mult-class data; this permits one, for example, to find class boundaries directly from sparse data (e.g., in image segmentation tasks) or to efficiently place centers for pattern classification (e.g., with local Gaussian classifiers). The same state variables can be used to define algorithms for determining adaptively the optimal number of centers for clouds of data with space-varying density. Some examples of the applicatin of these extensions are also given. |
| first_indexed | 2024-09-23T15:10:52Z |
| id | mit-1721.1/6613 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T15:10:52Z |
| publishDate | 2004 |
| record_format | dspace |
| spelling | mit-1721.1/66132019-04-12T08:31:38Z Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification Marroquin, Jose L. Girosi, Federico In this paper we present some extensions to the k-means algorithm for vector quantization that permit its efficient use in image segmentation and pattern classification tasks. It is shown that by introducing state variables that correspond to certain statistics of the dynamic behavior of the algorithm, it is possible to find the representative centers fo the lower dimensional maniforlds that define the boundaries between classes, for clouds of multi-dimensional, mult-class data; this permits one, for example, to find class boundaries directly from sparse data (e.g., in image segmentation tasks) or to efficiently place centers for pattern classification (e.g., with local Gaussian classifiers). The same state variables can be used to define algorithms for determining adaptively the optimal number of centers for clouds of data with space-varying density. Some examples of the applicatin of these extensions are also given. 2004-10-08T20:34:28Z 2004-10-08T20:34:28Z 1993-01-01 AIM-1390 http://hdl.handle.net/1721.1/6613 en_US AIM-1390 95179 bytes 1158231 bytes application/octet-stream application/pdf application/octet-stream application/pdf |
| spellingShingle | Marroquin, Jose L. Girosi, Federico Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title | Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title_full | Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title_fullStr | Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title_full_unstemmed | Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title_short | Some Extensions of the K-Means Algorithm for Image Segmentation and Pattern Classification |
| title_sort | some extensions of the k means algorithm for image segmentation and pattern classification |
| url | http://hdl.handle.net/1721.1/6613 |
| work_keys_str_mv | AT marroquinjosel someextensionsofthekmeansalgorithmforimagesegmentationandpatternclassification AT girosifederico someextensionsofthekmeansalgorithmforimagesegmentationandpatternclassification |