A recent conversation with a friend got me thinking of the intersection between technology, design, the preservation and flourishing of traditional handicrafts, and communities.
The Indian handicraft industry is a highly labor intensive one, with more than 7 million artisans, a majority of whom are women and largely underprivileged.
This industry, which is traditionally a major source of revenue generation in rural India, has been in decline (though there have been several efforts to support it), and has been hit hard by the pandemic as well.
What are the glaring gaps in the market for traditional craft? (specific to India, but this could apply to the world as well). To my mind the key gaps are in design, and in business building capacities.
Local artisans lack the ability to meet the needs of new markets and are forced to find low unskilled employment in urban industries. One of the major factors contributing to this is that artisans are not trained to contemporize their designs.
In this article, I’d like to focus on design and the role technology can play in meeting the current gaps.
While some work has been done on modernizing design, a lot of craft continues to center around traditional design, often not appealing to modern sensibilities, and thus not being able to build the foundation of a sustainable business. How can technology help? For example, AI techniques have been leveraged for emulating creativity and imagination – for image generation, style-transfer, image-to-image translation; for pattern generation, and color-transfer etc.
An interesting study (Raviprakash et al., May 2019) describes how AI techniques can be used to contemporize design, while keeping the underlying technique unchanged. It generated colored motifs and patterns that can be manufactured into physical products. This study experimented with using AI on the popular IKAT weave. Unlike other dyeing techniques, in IKAT the yarn is dyed BEFORE it is woven. This is what gives it its unique shading effect. This property was harnessed by the researchers to create a contemporary design.
The researchers first used a black motif using an AI technique trained on a set of 1000 paintings from a famous European painter, Piet Mondrian, and their gray-scale counterparts. The simplicity of these paintings along with the use of only primitive colors made them an ideal choice for our approach, since the model is able to learn primitive colorization of a motif from a relatively small training dataset.
The model used a generator which colorizes the input and a discriminator that learns to distinguish between the real paintings and the colorized images. The discriminator’s output determines the loss of the generator, which the generator tries to minimize, effectively colorizing images to make them indistinguishable from real paintings.
These motifs were re-colored with colors of an inspiration image using a statistical approach of global color transformation, and the design was post-processed to a grid that could be readily used for dyeing, as each cell is of a single color.
Products manufactured with designs generated using the above approach are found to be much more visually appealing than their traditional counterparts in the present market. Local artisans used these designs to manufacture and sell products successfully.
There are several such examples of how technology can modernize craft without compromising on the underlying uniqueness of a particular craft technique.
Investments need to be made in building such design capacity amongst artisans so they can once again take their place as valued centers of their communities.