Your bespoke AI image or video analysis model

The Guillin Group, Europe's leading manufacturer of food packaging, maintains a catalogue of nearly 25 000 references that its sales representatives need to quickly identify for customers. Galadrim trained the vision model that recognises packaging in a photograph and associates it with its reference, following a collection of training images conducted with the group's teams. It is now enough to photograph a product to immediately obtain its reference.

Castrol, a subsidiary of the BP group, operates an automated filling line where entering during the cycle exposes one to a serious accident, on an entirely offline site. We trained a detection model using several hours of video and 500 annotated images, chose the equipment capable of running it on-site, then installed the system on the line. A camera continuously monitors the dangerous zone: as soon as a person enters it, the alert goes off immediately, and no image leaves the factory.

Mecaprotec treats and finishes the surfaces of parts for the aeronautical industry, including paints and chemical treatments. Counting parts at the inspection station occupied its operators with a repetitive task, where the pace directly impacts productivity. Galadrim trained a computer vision model that detects and counts parts from station images, dimensioned the necessary equipment, then delivered the application through which quality teams monitor inspections on the pilot site.

Chanel's neuroscience division sought to understand the actual use of its products from filmed tests, with the aim of obtaining quantifiable data on how a person applies a cosmetic. Galadrim developed image recognition algorithms that track the user's head, hand, and fingers throughout the video. Each gesture is counted and classified by nature, whether a semi-circular massage or a tapping motion, which gives Chanel an objective measure of how a cosmetic is applied.

Avranches Automatic opens arcade halls in train stations, notably in Lyon Part-Dieu, and wanted to link each played game to its player's account without modifying the machines. Galadrim developed the vision module that continuously reads the video feed from each machine's screen, identifies the end-of-game score, and transmits it to the server from a card installed in the machine. A player unlocks a machine with their contactless card, plays, then finds their score and ranking in the mobile application delivered with the system.

YourArt brings together a media outlet on artistic news and a marketplace where artists and galleries exhibit and sell their works, with over a million visitors per month. Its visitors looked for artworks without mastering the specialist vocabulary of style, technique, or movement. Galadrim integrated two developers and a designer into the product team for two and a half years and developed the search feature that interprets a description in plain language as a reference image, based on the visual characteristics extracted from each work in the catalogue.

Gozoki, a French food group specialising in high-end ready meals, is deploying a network of continuously open autonomous stores, where the control of in-store operations relies entirely on automatic surveillance. Galadrim developed the vision system that analyses the store's cameras in real-time, tracks each person and product, and identifies situations to be reported. The web application delivered with the system displays the store's status in real-time and alerts security personnel to incidents.