1: \begin{abstract}

2: \baselineskip 16pt

3: We consider in this paper a contamined regression model  where the  distribution of the contaminating component

4: is known when  the Euclidean parameters of the regression model, the noise distribution, the  contamination ratio

5: and the distribution of the design data are unknown. Our model is said to be semiparametric in the sense that  the probability

6: density function (pdf) of the noise involved in the regression model is not supposed to belong to a parametric density family. 

7: When the pdf's of the noise and the contaminating  phenomenon are supposed to be symmetric about zero, we propose

8: an estimator of the various (Euclidean and functionnal) parameters  of the model, and prove under mild conditions

9: its convergence. We prove in particular that, under technical conditions all satisfied in the Gaussian case,  the Euclidean part of the model is estimated at the rate $o_{a.s}(n^{-1/4+\gamma})$,  $\gamma>0$.

10: We recall that, as it is pointed out in  Bordes and Vandekerkhove \cite{BV10},  this result cannot be ignored to go further in the  asymptotic theory for this class

11: of  models.  Finally the implementation and numerical performances of our method are discussed on several toy examples.

12: \end{abstract}

13: