Transforming retail: Unleashing the potential of foundation models
At YOLO Vision 2023, Intuitivo's José Benítez Genes showed how autonomous point-of-purchase machines use computer vision and foundation models in retail.
Nuvola Ladi2 min readListen to this article

Many retail innovations aim to bridge the physical and digital worlds.
This was on display at the YOLO Vision 2023 (YV23) event hosted by Ultralytics. At the Google for Startups Campus in Madrid, José Benítez Genes, Founder and Director of AI at Intuitivo, took the stage to present the company's Autonomous Point Of Purchase machines (APOPs), powered by computer vision.
Empowering retail transformation#
Intuitivo aims to transform retail by bringing digital technologies into physical shopping experiences. Through their APOPs, they aim to transform the traditional retail landscape, offering consumers unique and interactive shopping encounters.
But what exactly are APOPs? Essentially, they are autonomous machines equipped with cameras and driven by computer vision algorithms. Within 15 minutes, these machines can be integrated into existing retail setups, such as coolers, enabling autonomous transactions and redefining the shopping journey.
Navigating the realm of computer vision#
Central to Intuitivo's innovation is their advanced computer vision system, tailored to tackle the intricacies of real-world retail transactions.
During the talk, José provided a demonstration of their system in action, highlighting the challenges posed by complex interactions and occlusions.
In the world of retail, transactions are rarely straightforward. From obscured views to multiple interacting objects, Intuitivo's computer vision system faces many challenges. Through an end-to-end pipeline approach, the team works to resolve these complexities and track transactions accurately.
Scaling data processing with machine learning#
José also shed light on the task of scaling data processing with Intuitivo, tackling over 1 million images a day.
Using foundation models and techniques like knowledge distillation, the team has made its labeling process 60 times faster than traditional methods.

Fig 1. José Benítez Genes presenting at YV23 in Madrid at the Google for Startups Campus.
Embracing questions and charting future paths#
Audience members asked about APOPs' advantages over traditional vending machines, privacy safeguards, and Intuitivo's journey from Argentina to global expansion. José outlined the APOPs' cost-effectiveness, maintenance simplicity, and connectivity advantages, setting them apart in the retail landscape.
Privacy concerns were also addressed, with Intuitivo implementing stringent practices such as blurring faces in processed videos.
José also shared insights into Intuitivo's growth trajectory, which includes leveraging advancements in computer vision to further enhance their product offerings.
Wrapping up#
As José concluded his talk, the audience was left inspired by Intuitivo's vision and determination to redefine retail experiences through technology.
To learn more, watch the full talk.
Written by
Comments
Ask a question, share your results, or help someone else.










