In the nonlinear filtering model with signal and observation noise independ
ent, we show that the filter depends continuously on the law of the signal.
We do not assume that the signal process is Markov and prove the result un
der minimal integrability conditions. The analysis is based on expressing t
he nonlinear filter as a Wiener functional via the Kallianpur-Striebel Baye
s formula. (C) 1999 Elsevier Science B.V. All rights reserved.