Chanel
Analysing cosmetic gestures using AI to objectify customer experience.
The need
- The neuroscience hub Chanel's wanted to gain a scientific understanding of its product usage, with the help of video tests.
- The objective was to obtain quantifiable data how a person applies a cosmetic product to assess its quality.
- Generally, a low number of movements during product application is associated with ease of use of the product.
Galadrim's solution
- We implemented algorithms for video recognition in Python with the MediaPipe library.
- Our solution identifies the head, the hand and the fingers of the person applying the product and records their movements.
- We use artificial intelligence to count and categorise movements and understand their nature (half-circle massage, tapping...).
