Luxury cosmetics

Skin scoring and analysis

A major player in luxury cosmetics had developed a skin scoring algorithm. Problem: no one could verify what the AI was based on to make its diagnosis. We built the visualization system that makes predictions transparent and deployable in stores and on mobile.

Problem

The client already had an algorithm capable of analyzing a face photo and predicting skin quality scores. But he couldn't show that the AI was based on the right areas of the face to make its predictions. Without this proof, it is impossible to deploy the tool in store in front of customers: a skin diagnosis that cannot be explained does not create trust, it creates doubt. The challenge was no longer the performance of the model, but its transparency and credibility with the end user.

Vue rapprochée d’une coupe transversale colorée d’une géode montrant des couches concentriques de minéraux en jaune, marron, rouge et vert.

Solution

What we built

We deployed 2 Data Scientists in Scrum mode to design a complete system for visualizing and explaining predictions.

Step 1 — Data preparation. Creation and cleaning of the data set, standardization and data augmentation on the fly to ensure the robustness of the results on a variety of faces and shooting conditions.

Step 2 — Explainability of the predictions. Development of an interpretability solution exploiting the internal layers of the algorithm to identify and visualize the precise pixels that contributed to each prediction. The system shows exactly which areas of the face influenced the score.

Step 3 — Advanced visualization by GAN. Use of a GAN modifying the age to zero to calculate a wrinkle differential, offering an intuitive visualization of skin condition. The results were refined using heatmap, alpha alteration and guided filtering techniques for an aesthetic result adapted to the customer experience in store.

Step 4 — Multi-platform deployment. Optimization and porting of the system to work on machines in stores and on mobile, with performance tests validated on both supports.

Projects in the same category

See all projects

Supply Chain Optimization Application

A pharmaceutical distribution player had to rethink its entire supply chain: pharmacy assortment, inventory management and delivery channels. Operational research alone was no longer enough. We built the application that transformed its operations.

Social media trend detection

In a market where consumer behaviors change faster than decision cycles, a customer needed to anticipate trends instead of experiencing them. We built the platform that turns social media noise into actionable signals.

Yield of agricultural fields

The climate subsidiary of a French insurance leader needed to predict field yields across Germany to price a new drought insurance offer. Internal data was not enough. We built the predictive models that made the product marketable.

Customer needs analysis platform

A large French group needed to understand the current and future needs of its customers by aggregating a massive volume of consumer data. After a POC of 8 weeks, we industrialized a complete platform deployed on all French entities and is now being extended at the group level.

Skin scoring and analysis

Beaconing detection

A large group's security team needed to identify beaconing signals, regular network traffic sent by potentially compromised machines to servers controlled by attackers. The volume of logs made manual detection impossible. We built the anomaly detection system capable of processing this data on a large scale.