AI in lipstick

From Sea Vision comes a visual inspection system dedicated to the most iconic cosmetic formulations.

In what will go down in history as the year Artificial Intelligence becomes widely available, SEA Vision introduces a-eye lipstick, a new AI-based lipstick inspection system.

The solution aims to improve the packaging of a product currently recognised as one of the world’s best-selling cosmetics. This goal was achieved by automating processes that, until now, were largely subject to human control.

Dealing with inconsistent standards

The numerous distinctive aesthetic features together with the complexity of the industrial process required to perfectly amalgamate the pigments, oils, waxes and emollients in its formula mean that lipstick is complicated to produce and to verify in terms of product quality. Even today, despite all the efforts made during production to prevent and control a whole series of potential defects, some flaws still pass through controls performed by pre-neural technology undetected. This leads to costly reprocessing or, in the worst-case scenario, to placing these sub-standard products on the market, posing serious risks for businesses in terms of brand reputation.

AI for higher quality

The use of AI sets out to remedy a large proportion of these potential defects. A-eye lipstick, developed by a team of experts from SEA Vision and its subsidiary ARGO Vision, is based on semantic segmentation of the lipstick areas in order to identify every possible flaw pixel-by-pixel. The system learns how to discern an ever-increasing variety of complex defects. Learning takes place both on the basis of proprietary datasets – a mix of real and synthetic images generated with the most advanced data augmentation and neural generation techniques – and by combining different parameters and patterns observed over time.

Machine learning applied to cosmetics

The a-eye lipstick technology aims to generalise the “concept” of lipstick, making the algorithms more and more specialised in quality control regardless of the product’s possible shape, colour and texture (e.g. satin, transparent, matt, pearl, glitter, etc.). As Alessandro Ferrari, CEO of ARGO Vision, explains:

“There is one aspect that distinguishes this innovation, differentiating it from conventional vision systems which can detect defects in images through comparison with models already known at the outset. It is about the system’s ability to learn from past examples, and thus constantly increasing its analysis capability, which is more or less what we humans do when we learn from experience.”