I suspect the problem you're running into is due to the fact that all objects you get from Poser are implemented in native code and that the memory is leaking because the underlying allocated memory can't be released by the Python garbage collector.

The only thing you can do is to try using del on large lists, tuples, and arrays allocated via numeric or numpy, to see if you can improve the situation.

It may also be the case that the use of native objects prevent the debugging tools from detecting where leaks are occuring.