What is sCa fUndamental hypothesis for SCa When the transformation dictionary is big enough compared with the signal size, there must be sparse expression for signal in this transformation domain
What is SCA ❑Fundamental hypothesis for SCA ➢ When the transformation dictionary is big enough compared with the signal size, there must be sparse expression for signal in this transformation domain；
What is sca Overcomplete dictionary 过完备库中的分量不一定满足正交性, 称为原子(aom) 原子个数远大于信号的维数; 在过完备库中,信号的表达是不唯一的 Underdetermined system)
What is SCA ⚫ Overcomplete dictionary ➢ 过完备库中的分量不一定满足正交性， 称为原子(atom)； ➢ 原子个数远大于信号的维数； ➢ 在过完备库中，信号的表达是不唯一的 (Underdetermined system)
What is sca o How to select and construct overcomplete dictionary Different overcomplete dictionary for different signal and purpose
What is SCA ⚫ How to select and construct overcomplete dictionary ➢ Different overcomplete dictionary for different signal and purpose
What is sca a Overcomplete dictionary kinds Pure overcomplete Mixed overcomplete Special overcomplete
What is SCA ❑ Overcomplete dictionary kinds ➢ Pure overcomplete ➢ Mixed overcomplete ➢ Special overcomplete ➢ ……
Approaches for SCA 概率框架模型( Probability Framework,PM Zibulevsky. et al. 200 >最佳正交基法( Best Orthogonal Basis, BOB Coifman, et al, 1992) Focuss算法( Gorodnitsky,eta1,1997) >拉格朗目乘子法 >匹配追踪法( Matching Pursuit,MP
Approaches for SCA ➢ 概率框架模型(Probability Framework，PM. Zibulevsky , et al,2001) ➢ 最佳正交基法(Best Orthogonal Basis,BOB. Coifman, et al ,1992) ➢ Focuss算法(Gorodnitsky ,et al,1997) ➢ 拉格朗日乘子法 ➢ 匹配追踪法(Matching Pursuit,MP)