Use of cheap equipment

A PowerGlove has extreme noise and accuracy limitations, as should be obvious by now. However, it is sometimes possible to use a small amount of information to extract enough identifying and discriminating attributes from data to uniquely identify it.

Also note that sign language is two-handed, but that we only have one glove. This is a significant problem, but its severity is alleviated by a number of factors:

• A very significant proportion of signs are one-handed (approximately 40 per cent).
• Of the signs that are two-handed or double-handed, a significant proportion are symmetrical or near-symmetrical.

• Of the remaining signs, the dominant hand usually makes the most complex movements, with enough character to identify it uniquely.

This does not mean, however, that there are no signs solely distinguished by what the passive hand is doing; it is just that fortunately, this does not occur in a large number of cases.

This will certainly be a challenge, but if it is to be an effective communication tool, that can be used in the general domain, then a large proportion of the population could not afford to purchase US$5000 gloves. At US$500 or below, however, people might be able to afford it, even if they have to put up with a slightly higher error rate.