Classical state space representations treat patterns as fixed points. But patterns like the face set of an individual are known to show continuous variability. Yet these are similar and tend to cluster together. Such similar patterns form a pipeline in state space that can be used for pattern classification. A learning algorithm to model the pipeline is presented in this paper. A least squares estimation approach that utilizes interdependency between points in training patterns is used to form the non-linear pipeline. Multiple patterns can be trained by having separate lines for each pattern. Points in each pattern are now projected onto the respective pipeline. Unlike most other manifold techniques, the proposed method provides an easy intuitive way to place new points onto the manifold. Given a test point∕face, the classification problem is now simplified to checking the nearest neighbors. This can be done by finding the minimum distance pipe-line from the test-point. The proposed representation of a face image results in improved accuracy when compared to the classical point representation.