Output details

Blind Separation of Image Sources via Adaptive Dictionary Learning

<12>This paper presents a breakthrough unifying algorithm which benefits from both sparsification and source separation to enable efficient separation of sources with potential sparsity in some unknown domain. This is the first time compressive sensing and blind source separation have been combined. Previous methods for sparse data, reliant on clustering, were not accurate. Data in nature is mixed and sparse, so this algorithm allows the exploitation of full properties of natural data such as images or biomedical data e.g. brain signals, or separation of maternal and foetal heartbeats.