PhotoEye AI designed to help companies find, filter and edit content with innovative assessment and optimization tools
December 9, 2019 – Dreamstime, the world’s largest community in stock photography, with 27 million users and over 118 million images, recently released its proprietary Artificial Intelligence technology (AI) to third party companies. Implemented on Dreamstime’s website in 2017, this technology uses sophisticated algorithms to screen and optimize images and uses machine learning to constantly adjust parameters to improve its capabilities. Dubbed PhotoEye, this customizable solution offers users content filtering, copyright and logo detection, model release count, automatic keyword suggestion, smart cropping, privacy tools and scalability to improve results for companies and their customers.
Dreamstime’s AI offers an advantage over generic models as it learns based on feedback from human photo editors, giving it an edge as it is directly tailored to visual content. This feedback reduces the number of false positives flagged in the screening process and overall improves the tool’s performance.
“We are constantly innovating and seeking new ways to contribute to the improvement and growth to the industry. AI is not just the latest technological fad; it has real potential to shape the future of just about every aspect of life, and it has immense potential especially in stock photography,” said Serban Enache, CEO and cofounder of Dreamstime. “We are releasing this tool for other companies to use so they can reap the same benefits that this technology has provided Dreamstime.”
Companies can save time by implementing this advanced photography solution and consequently reducing the hours spent by humans on mundane, manual tasks. With a more efficient search and editing process, this tool has the capability to improve marketing, advertising and creative campaigns and projects through several helpful features. For these reasons and more, the two Fortune 500 large tech companies Dreamstime currently has broad partnerships with are considering incorporating this PhotoEye as a plugin.
PhotoEye features include, but are not limited to:
● Content Filter: To reduce the search time and find only the most relevant images, PhotoEye filters and flags any sensitive media, with a machine learning-based engine that is eight times more accurate at filtering potentially offensive content, including nudity, substance abuse and violence, than other detection tools, such as Google Vision. PhotoEye provides a 61% match as opposed to only a 20% match by Google vision, as tested by our editors and AI engineers.
● Copyright and Logo Detection: To avoid copyright infringement, PhotoEye will also detect and copyrighted logos or brands that appear in photos or videos, no matter the scale.
● Model Release Count: PhotoEye AI will organize and optimize model-filtered searches by associating model releases and detecting people in your visual content.
● Keywords and Categories: This tool has also learned from billions of images and keywords, providing automatic keyword suggestions to tag your content quickly and efficiently, saving time and improving your library and SEO.
● Privacy: PhotoEye was developed with user privacy top of mind. All content filtered through PhotoEye is for PhotoEye only; images and videos are evaluated and optimized through an objective lens as any information collected by the tool is anonymized and not traceable – while PhotoEye is fed by user data it will never be traced back to an individual user.
In the digital age where uploading and finding content is easier than ever, Dreamstime strives to provide innovative and user-friendly solutions to make users’ content stand out. To see the capabilities PhotoEye has to offer, visit https://www.dreamstime.com/photoeye for a free live demonstration. For more information about Dreamstime, visit https://www.dreamstime.com/.