1: \begin{abstract}                % Abstract of not more than 250 words.

2: % We prove the existence, on a given set and under some conditions, of a nonlinear Luenberger observer. 

3: KKL (Kazantzis-Kravaris/Luenberger) observers are based on the idea of immersing a given nonlinear system into a target system that is a \textit{linear} stable filter of the measured output.

4: In the present paper, we extend this theory by allowing this target system to be a \textit{nonlinear} contracting filter of the output.

5: We prove, under a differential observability condition, the existence of these new KKL observers.

6: We motivate their introduction by showing numerically the possibility of combining convergence speed and robustness to noise, unlike what is known for linear filtering.

7: % Instead of immersing the dynamical system into a linear filter, we use a nonlinear one, which can have some interesting properties in terms of speed, precision and robustness to input noise. The proof is made under the assumption of a high-gain filter.

8: \end{abstract}

9: