What is Vision AI?

Vision AI protects a product by a single photo taken with a mobile or standard industrial camera, either during the manufacturing process or as the final product. This photo serves as the product's unique identifier, functioning just like a human fingerprint. This identifier, called FeaturePrint™, is used to link the product to its digital identity.

What are the main use cases Vision AI is covering? 

The key use cases of Vision AI are:

• Supply chain tracking (B2B) -  Matches individual products to uniquely numbered certificates of authenticity

• Identifying fake returns (B2B) - Verifies if products returned for service, warranty, or refund are original

• Product authentication (B2C) -  Provides consumer access to product authenticity

• Enhancing security in DPP (B2C) - DPP itself does not guarantee authenticity; Vision AI can ensure individual product authenticity.

• Connecting physical and digital product (B2C) - Useful especially where QR codes or NFC chips are not applicable (e.g., watches, jewellery, eyewear)

How is FeaturePrint™ different from other identifiers?
FeaturePrint™ uses patented algorithms to uniquely serialise items. Unlike traditional image comparison, it employs AI to create 20,000 reference points within a captured image, combining these unique references for identification. FeaturePrint™ allows 1:1 verification in near real-time through ensuring an acceptable number of reference points.

Does Vision AI require special devices or additives?
Vision AI requires no additives or external markers. It operates without the need for barcodes, QR codes, stickers, tags, 'invisible' dust, markers, holograms, special cameras, or special lighting.

What range of products can Vision AI protect?

Vision AI has successfully worked with leather goods (bags, wallets, belts) and ready-to-wear items (including woven materials and silks), and has expanded its capabilities to include shoes, jewelry, sunglasses, and watches. 

Currently, Authentique does not support products that can easily change their structure, such as liquids, powders, or feathers. However, Authentique can protect seals and packaging, such as those on perfume bottles and high-value wines and spirits.

Can Vision AI protect rings and small metal objects?  What is the minimum dimension required?

Vision AI has been successfully tested on jewellery and accessories, including objects as small as the head of a screw (minimum 0.1 cm radius).

What about wear-and-tear? Can it stop working?
Vision AI has been tested for wear-and-tear across a wide range of products. During the initial phase of the project, our team collaborates with brands to define expectations and establish parameters. We also offer a wear-and-tear program for brands to test the technology's durability.

In rare instances where a fingerprint is heavily damaged, such as when a product is completely sandpapered, identification may become impossible. In such cases, We recommend brands to utilise consumer touchpoints to re-register fingerprints. For instances of watches, re-registration can occur during overhauls, polishing, or repairs.

How does light and reflection affect FeaturePrint™?
FeaturePrint™ does not require special lighting for registration and verification. However, it is recommended to have control over the external environment during these processes. Avoid direct sunlight and minimise heavy reflections to ensure optimal results.

How well does Vision AI fit into the manufacturing process? Is it scalable? 

Vision AI can be registered either during the manufacturing process or as the final product. Authentique provides expert guidance on seamlessly integrating Vision AI into brands’ manufacturing process. For smaller-scale products, registration can be done manually using mobile phone cameras. For larger-scale products, registration can be automated using industrial cameras.

How much does it cost?

Authentique offers a wide range of services tailored to each brand's specific needs. For any inquiries, please contact us at contact@ordre.com