LOS ANGELES, CA - JANUARY 06: Judy Greer attends FIJI Water at the 76th Annual Golden Globe Awards on January 6, 2019 at the Beverly Hilton in Los Angeles, California. (Photo by Stefanie Keenan/Getty Images for FIJI Water) [01 modeling photography]
Model,Jessica Serfaty spends time with her friends in Miami Beach,Florida<P>Pictured: Jessica Serfaty<B>Ref: SPL1188923 011215 </B><BR/>Picture by: Robert O'neil/Splash News<BR/></P><P><B>Splash News and Pictures</B><BR/>Los Angeles:310-821-2666<BR/>New York:212-619-2666<BR/>London:870-934-2666<BR/>email@example.com<BR/></P> [01 modeling photography]
Stephon The Surgeon Morris cuts off his hand wraps after his boxing session at UMAR Boxing Gym Baltimore, Md., June 18, 2014. Stephon is a 21-year old heavyweight boxer from Baltimore who has been training at UMAR for three years and is going to punch his pro card later on this year. (USAF photo by Tech. Sgt. Andy M. Kin) [gym stock photography]
To use all the features of this site you must be logged in. If you don't have an account you can sign up right now.
AI Photo Analysis
Evaluation of picture quality and feel has been a long-standing issue in photo handling. While specialized quality appraisal manages to measure pixel-level debasements, for example, commotion, obscure, pressure antiques, and so forth., tasteful evaluation catches semantic level attributes related to feelings and excellence in pictures. Hossein Talebi, Google Software Engineer said, “Our proposed network can be used to not only score images reliably and with high correlation to human perception, but also it is useful for a variety of labor-intensive and subjective tasks such as intelligent photo editing, optimizing visual quality for increased user engagement, or minimizing perceived visual errors in an imaging pipeline.”
The AI photo tagging engine are automaticaly annotating photos with multiple image tags, to enhance the quality of visual representation of the trained CNN model. It is based on a large-scale multi-label image database with 18M images and 11K categories. The AI Assessment Rating are trained using two models (AVA & TID2013) to predict the aesthetic and technical quality of photos. The models are trained via transfer learning, where ImageNet pre-trained CNNs are used and fine-tuned for the classification task. AI photo processing are running in the background using several GPU-powered backend servers. These AI technologies are based on Vedere AI Engine, you can check out the website at www.vedereai.com
AI Image Search
Full text search (e.g. 'delicious food') to search the images name and description.
Minus symbol (e.g. '-beach') to remove specific words from images name and description.
filetype:extension (e.g. 'filetype:png') to search specific images filetype.
camera:brand (e.g. 'camera:nikon') to search specific brand of camera used.
iso:speed(e.g. 'iso:1250') to search for images taken using specific ISO speed.
f:number (e.g. 'f:6.3') to search for images taken using specific aperture.
mm:number (e.g. 'mm:50') to search for images taken using specific focal length.
tags:name (e.g. 'tags:valley') to search specific image classifications detected by AI.
caption:name (e.g. 'caption:blue') to search specific wording in image captioning detected by AI.
category:name (e.g. 'category:macro') to search specific photo category.
faces:number(e.g. 'faces:3') to search for images with the number of faces.
Combining any of the above (e.g. 'caption:blue tags:gown category:fashion') to search for 'blue' in AI image captioning that is a 'gown' in AI image tagging under the 'fashion' category.
Another example (e.g. 'iso:100 camera:canon tags:bikini') to search for 'canon' camera used together with ISO setting of 100 and contains bikini in the AI image classification.