How to decide which lipstick to buy without physically trying
Introduction
VIrtual Makeup Tryon and Product Recommendation is cutting edge use case of Augmented reality in the fashion e-commerce industry.
Deciding which Lipstick shade suits best based on AI recommendation for individual skin tone can make women look beautiful and confident about their choice. It will also help e-commerce companies reduce costs associated with product returns.
It can be used in multiple industries where virtual tryon can help people choose products.
The frictionless shopping experience is very important in a cut-throat competition to stand out of the crowd and the feeling of ‘Try Before You Buy will dismiss any hesitation against e-commerce purchase
What is Augmented Reality Beautification and Why Does it Matter?
Now, you should have a fair idea of what we are going to explore further in this blog post It’s about technology that will revolutionize how fashion/cosmetics products are sold.
The beauty industry is soaring to $530 billion and growing faster, the shopping experience is a prime factor which generates brand loyalty and drives sales.
How Sephora did it?
Here we are going to briefly talk about Sephora (a French cosmetics major) that deployed a virtual assistant in a web application that list products that Sephora manufactures.
For Example, Lipstick Try On Page list of all available shades of lipstick where user can select a particular color and apply live on webcam stream of her face
Solution design briefly includes:
- Augmented Reality for face-tracking>
- ML based face feature detection
- OpenCV / Image Processing to produce overlay after product application
- Rendering /Player for the user to see
Modiface – In-store Mirrors
Based on the same solution stack but using standalone Hardware/Kiosk in a physical location to drive sales.
Like every product, there are some features that were spot on while some were nowhere close to reality. After evaluating Sephora virtual assistant, here’s what we found:
Dazzling Features
- Eyelashes – It tracked perfectly and worked exactly as expected in the real world
- Eyebrows – Just like eyelashes, eyebrows simulation will also look great
- Lipstick – Lip Color try on was good but for a few of the darker shades, lighting wasn’t accurate
Dimmed Show
- Blush – skin color correction was not tracked accurately
- Foundation – lightening/difference was not noticeable enough
Limitations
- Image segmentation is still in the infancy stage. Hence, applying hair color doesn’t work with accuracy.
- Foundation/Blush depends on detecting skin surface accurately with the color application, which is nowhere close to reality.
With advancements in detection/Tracking technology and new algorithms,We expect feature improvements in future releases and many more use cases will be discovered. At the current stage apart from a few use cases, Augmented reality is still in its infancy and needs more research & development efforts before it becomes a part of everyday life.
Let us know which of our services you’re interested in, and our team will get back to you shortly!