The claim that one looks at women in a purely esthetic way can be challenged, see Kant's distinction between free and adherent beauty, in SEP, "Kant's esthetics and teleology".
A suitable test could be : "Would I get as much aesthetic pleasure in case I knew this person only appears as if they were a woman I desire?". Also, occording to Kant, this could be a test : " Would I get so much aesthetic pleasure in case I knew this person is a pure hologram, not representing a person really existing".
In order a judgment to be purely esthetic, the existence of the object should not be taken into account; this is a condition for the judgment to be free of all interest ( that is of all need or desire).
This article so far has been concerned only with “pure” judgments of
beauty. But Kant also allows for judgments of beauty which fall short
of being pure. Judgments of beauty can fail to be pure in two ways.
(a) They can be influenced by the object's sensory or emotional
appeal, that is, they can involve “charm” [Reiz] or emotion [Rührung]
(§13). (b) They can be contingent on a certain concept's applying to
the object, so that the object is judged, not as beautiful tout court,
but as beautiful qua belonging to this or that kind. The second kind
of impurity is discussed in §16 in connection with a distinction
between “free” [frei] beauty and “adherent” or “dependent” [anhängend]
One reason to think that the distinction is important is that Kant
seems to suggest that all judgments of beauty about representational
art are judgments of adherent rather than of free beauty, and hence
that they are all impure. While some art works can be “free beauties,”
the examples Kant gives are all of non-representational art: “designs
a la grecque, foliage for borders or on wallpaper…fantasias in music,”
and indeed, Kant adds, all music without a text (§16, 229). It might
be supposed from this that Kant's core account of judgments of beauty
is only peripherally applicable to art, which would make it largely
irrelevant to the concerns of contemporary aesthetics. However, this
consequence is debatable. For example, Allison argues that judgments
of adherent beauty contain, as a component, a pure judgment of beauty.
The purity of this core judgment is not undermined by its figuring in
a more complex evaluation which takes into account the object's
falling under a concept (2001, pp. 140–141).
Kant's suggestion that representational art has “adherent” rather than
“free” beauty, and that judgments about such art fail to be pure,
might also invite the objection that Kant takes nonrepresentational
art to be superior to representational art, so that, say, wallpaper
designs are aesthetically more valuable than the ceiling of the
Sistine Chapel. This objection is challenged by Schaper (1979, ch. 4,
reprinted in Guyer 2003) and by Guyer (2005a, chs. 4 and 5).
Further discussions of the distinction between free and adherent
beauty include Scarre (1981), Lorand (1989), Gammon (1999), Kalar
(2006), pp. 82–89, and Zuckert (2007), pp. 202–212.