Enhancing Just Walk Out technology with multi-modal AI
Machine Learning Blog
This article discusses the latest advancements in Amazon's Just Walk Out technology, which allows customers to shop without needing to check out. It introduces a new multi-modal AI system that enhances the accuracy and scalability of the technology.
Specifically, the article covers:
- The challenges of handling complex shopping scenarios with previous versions of Just Walk Out technology
- The solution: a new multi-modal foundation model (FM) designed specifically for retail environments
- Key features of the multi-modal FM, including flexible data inputs, AI tokens for representing shopper journeys, and continuously updating receipts
- The training process for the FM, involving auxiliary tasks, data flywheel, pre-training, and fine-tuning
- The benefits of the new system, such as improved accuracy, scalability, and adaptability to new store formats and customer behaviors
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
Nov 13
2024
2024
Create frictionless retail experiences with Just Walk Out RFID lanes
Jul 10
2024
2024
Enabling Credit Card Entry with Cognizant Customer Concierge Solution for Amazon Just Walk Out Technology
Jan 10
2025
2025
New shopper authentication experiences from Amazon Just Walk Out
Jan 4
2024
2024
Cognizant Customer Concierge Solution Connects Amazon’s Just Walk Out Technology to Any Retail System
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.