Amazon is upgrading its AI model to improve accuracy.
In a blog post on Wednesday, 31 July, the mentioned e-commerce giant, which also owns the best American delivery service, said the new multimodal AI foundation model uses advanced machine learning techniques similar to those applied to digital products powered by generative artificial intelligence. In this case, it analyzes data from cameras and sensors simultaneously.
The AI model announced by Amazon on Wednesday is a sign that the e-commerce giant is moving away from its previous method of analyzing information sequentially to address issues like camera obstructions and complex shopping behavior.
The company is also currently planning to license its Just Walk Out checkout system to more stores.
Jon Jenkins, vice president of Just Walk Out technology, said that over the past couple of years, it has become clear that it is possible to combine all the different signals into a single model and train a machine learning system to operate with all of that information at once to produce a receipt. According to him, the initial experience of applying a multimodal approach using a single foundation model has shown that in this case, the receipt generation process is faster, more efficient, and more accurate.
Jon Jenkins says that the foundational technology allows Just Walk Out to process and integrate multiple types of data inputs, including video from cameras, weight signals from sensors, RFID signals, and images of items taken off shelves. According to him, the mentioned technology is gaining traction beyond its own stores, with over 170 installations in third-party locations.
