This section features a number of tutorials illustrating some of the main algorithms implemented in VLFeat. The tutorials can be roughly grouped into two categories. The first class of algorithms detect and describe image regions (features). The second class of algorithms cluster data.
Features
- Scale Invariant Feature Transform (SIFT). Getting started with this popular feature detector / descriptor.
- Maximally Stable Extremal Regions (MSER). Extracting MSERs from an image.
Clustering
- Integer optimized k-means (IKM). A quick overview of our fast k-means implementation.
- Hierarchical k-means (HIKM). Create a fast k-means tree for integer data.
- Agglomerative Information Bottleneck (AIB). Cluster discrete data based on the mutual information between the data and class labels.
- Quick shift. An introduction which shows how to create superpixels using this quick mode seeking method.
Other
- MATLAB Utilities. A list of useful MATLAB functions.