subspace clustering

For very high dimensional data, subspace clustering algorithms attempt to find subspaces underlying the data and then find clusters in each such subspace. The final clustering is obtained by combining the clusters in each subspace. Basically, the hope is that individual representations in each subspace will remove the overall clutter, and lead to clearer view of the data which will be easily clusterable.

What are some of the best known algorithms that have shown empirically good results?