Landing AI, Andrew Ng’s cloud-based platform for computer vision, has announced a free trial of its flagship CV offering, LandingLens. The company says the free trial offer is meant to democratize the creation of artificial intelligence among companies large and small.
“Anyone can now create an AI model to process images. This has immediate applications in automotive, electronics, biotech, pharmaceutical, agriculture, and other industries,” said Ng, Landing AI founder and CEO. “AI will create untold value for all industries and any company, with or without a machine learning team, can unlock the value of getting computers to see.”
LandingLens is a data-centric AI computer vision software platform that emphasizes data quality over quantity. Companies do not always have large datasets for many CV projects, and the LandingLens algorithm works even with limited training data. “LandingLens’ step-by-step user interface vastly simplifies the training and deploying of AI models, making the technology accessible to individual subject matter experts within companies,” the company said in a release.
(Source: Landing AI)
The platform improves data quality through an advanced labeling technology. Mislabeled images are automatically found and corrected. Consistency throughout the data is ensured by label book, a feature where collaborators can align on clear definitions of classes and labels. With this, multiple users can label images and work towards consensus. The platform continuously learns and updates the model by adding new data from the deployment environment to re-train the model.
Landing AI says over 1,000 customers have signed up to use LandingLens in recent months. In addition to this trial offer, there is also a new pricing schedule that allows users to pay as they go should they decide to continue beyond the trial period.
One LandingLens customer is OmniAb, an antibody discovery company leveraging computational, hardware-based, and genetic technologies to enable rapid development of therapeutics. Landing AI says the company used LandingLens to automate its visual inspection process, leading to a throughput of 10x and significantly increased efficiency.
A project screen from LandingLens. (Source: Landing AI)
“Because of the throughput of our platform, we are able to image each of the millions of cells we screen. Even in highly complex campaigns, we find the rare samples. Out of a thousand images, we may only have maybe 10 images that have what we are looking for,” said Bob Chen, PhD., OmniAb senior director of system engineering.
Ng has long been a proponent of data-centric AI, emphasizing data quality for AI projects. This democratized approach allows companies with limited data to create AI products without the need for massive datasets and computing infrastructure.
Watch a video from Ng on how to create a CV project and find a link to sign up for the free trial here.
Related Items:
How Data-Centric AI Bolsters Deep Learning for the Small-Data Masses
Andrew Ng’s Computer Vision Startup Nabs $57 Million
The Key to Computer Vision-Driven AI Is a Robust Data Infrastructure
