[ml] WiiMote sync-up
trochee at gmail.com
Tue Jun 16 22:23:47 UTC 2009
Josh Myer wrote:
> There was a quick aside on how to smooth/sample that data: my thought
> is to place datapoints into buckets of a fixed width, and take the
> midpoint of each bucket. This requires taking the mean of the samples
> within each bucket. Then, to smooth in missing buckets, we do linear
> interpolation. The alternative is to take sample points at the edge
> of each bucket, with some form of interpolation. My gut feeling is
> that we want to take means because the sensors are noisy, and making
> use of as many samples as possible is beneficial. We'll see how this
> works out as users change direction (ie: an up-to-right transition
> might fare poorly with this sampling strategy).
smoothing is a good idea. my experience with speech suggests that you
can smooth better doing a sliding window, rather than bucketing --
unless I interpret your use of 'bucket' wrong. the same sample may
belong to multiple frames, if that makes sense.
the window need not be rectangular; under some circumstances a
triangular window or even a sliding Gaussian makes more sense, depending
on the properties you're trying to capture.
just my $0.02, from a kibitzer who has other Wednesday night regularly
More information about the ml