Technoculture of Handcraft: Fine Gesture Recognition for Haute Couture Skills Preservation and Transfer in Italy

In many different fields specialized artisans have become difficult to find, as their knowledge and their practical abilities are anything but easy to teach and transfer with training models that are no longer based on traditional master-apprentice relationships. This situation is particularly exacerbated in all those high-end haute couture companies that have built their glory on their style, as well as on their highly specialized craftsmen capable of turning leather, wool and other materials into inestimably valued shoes, bags and clothes. Today, alternative solutions can be found as technology can be put to good use to encode, and thus preserve, all this expertise, providing digital means of passing it on to new generations. Modern technologies have already been experimented with manual craftsmanship, for example, in the context of knitting [Rosner and Ryokai, 2009], although focusing most on its amusement and social aspects rather than on its knowledge encoding ones. Tracking techniques could support, instead, a system capable of digitizing the hand gestures performed by an artisan while handcrafting. However, such proposals very often require users to wear specific garments, not always suiting the scenario, as artisans reluctantly bear the use of invasive modern technologies. We show that noninvasive technologies can be exploited to encode and thus preserve artisanal knowledge by presenting a system based on a set of fine gesture recognition algorithms, as derived by [Roccetti et al., 2010], that require no peculiar attire as they solely utilize a frontal webcam, positioned at a close distance from a handcrafter. We witness the viability of using such system for tracking a handcrafter in two important phases of shoe making: hammering and sewing a shoe.