Dear colleagues, Please find below the abstract of next seminar. It will be presented by Jun Li. Best regards, Jun Li ---------------------------------------------------------------------------------------- MINIMUM VOLUME SIMPLEX ANALYSIS: A FAST ALGORITHM TO UNMIX HYPERSPECTRAL DATA I will present the minimum volume simplex analysis (MVSA) algorithm for hyperspectral unmixing, which is a continuous work with what presented by Professor Bioucas in last seminar. The new MVSA algorithm here solves a linear problem other than a sequencial of quadratic programmings. The complexity of the new algorithm is O(m^2) much faster than the original method. We illustrate the state-of-the-art performance of the MVSA algorithm in unmixing simulated data sets with no pure pixels. The new algorithm performs better than or comparable to MVES and almost the same as the old MVSA algorithm; yet it has a computational complexity much lower than those algorithms.