Box Taps Google's Cloud Vision Technology For Image Recognition

The Google Cloud Vision Integration will allow customers to better manage images uploaded to Box and to extract more value from them, according to Ben Kus, Box senior director of product management.

Box Google Image Recognition

Box will use Google's Cloud Vision technology to help its customers manage image files uploaded to its content management platform.

The integration will enable enterprises to automatically catalog images uploaded to Box into thousands of categories that make them easier to identify, organize, manage and search.

"Images are the second most common and fastest growing content type in Box," said Ben Kus, senior director of product management at Box in a blog. Enterprises of all sizes are uploading massive libraries of images that include product photos, completed forms captured via mobile phones, images of structures and buildings as well as images of claims forms and loan applications.

"To help unlock the value of these images to your business, we're exited to introduce image recognition with Box," Kus said.

Google's Cloud Vision API lets organizations enable vision detection features such as labeling, face or landmark detection and content tagging, within applications.

The technology for instance can enable capabilities for classifying images into thousands of categories like buildings, animals, landmarks, humans, flowers and transportation and pictures.

It can be used to detect and index words, objects and faces within an image and lets organizations build metadata on their image catalog. Cloud Vision uses machine-learning technology to improve upon its image recognition capabilities.

The integration of Cloud Vision into Box means that when any image is uploaded to Box, any objects and typed text or handwritten notes in images will be automated detected and indexed as metadata. This adds actionable context to the images, Kus said.

Beta customers that have tried the new image recognition capabilities in Box have already benefited from it, he said.

For example, one media company has been using the feature to automatically tag large volumes of photos from freelance photographers so they are able to find and use them more efficiently.

A retail organization is using image recognition to manage the company's massive collection of product-related photos while a global real estate firm is leveraging Cloud Vision's optical character recognition capabilities to enable digital workflows for paper based agreements and leases, Kus said. This has enabled employees to bypass the manual tagging process for classifying assets that was used previously, he noted.

The Cloud Vision integration is part of a broader effort at Box to leverage machine-learning technologies to help companies manage content and extract more value from it Kus said.

Google has been using Cloud Vision and similar technologies internally for several years. The company released Cloud Vision API in February 2016 as part of an effort to help developers and businesses embed enable image recognition in their applications and services.

Google currently allows pretty much anyone that wants their images analyzed to submit them via the Cloud Vision API. Google does not charge for the first 1,000 images that are uploaded to the API each month. After that pricing for the service is based on the specific tasks for which customers use the API.

For example, the cost for label and text detection within images is $1.50 for between 1,001 and 5 million images per month and $1 for between 5 million and 20 million images per month. Prices for services like document text detection in images are higher at $3.50 for between 1,001 and 5 million units per month.

Jaikumar Vijayan

Jaikumar Vijayan

Vijayan is an award-winning independent journalist and tech content creation specialist covering data security and privacy, business intelligence, big data and data analytics.