Visual Search

VisualSearch Chrome Plug-In

Big Picture Analytics

At ImageVision we see the Big Picture through a powerful set of image analysis tools that can decompose digital imagery into discreet component attributes enabling advanced recognition and detection capabilities. Big Data is great for numerical and textual data but what about the explosion of visual content?

With over 3+ Billion photos and videos shared via social media every day, arguably social media’s greatest asset is its free supply of user-generated and user-shared content.

By combining real-time visual recognition technology with artificial intelligence and machine learning, ImageVision provides social media sites the tools to automate the recognition and monetization of their visual content to both grow revenue and protect advertiser brands.

AD Personalization

ImageVision visual analytics solutions, help advertisers and publishers improve ad relevancy and click through rates (CTR) by targeting ads based upon user Interest, leading to a better consumer experience.

ImageVision’s technology recognizes objects, and patterns in images and videos, actually learning consumer current preferences and interests.

Consumers can opt-in to services embedded in mobile apps to both help manage image content and personalize the relationship with the brands they love. For them, the power of image analytics automatically organizes visual content, managing sharing and archival across devices, social media and storage services. Simultaneously, advertisers and publishers improve ad relevancy and click through by targeting ads based upon user interest, leading to an overall better consumer experience.

ImageVision’s technology recognizes objects, and patterns in images and videos, actually learning consumer current preferences and interests.

The Cookie Crumbles

Based on a recent survey conducted by Harris Interactive titled 2013 Bad Ads Survey, “87% of Americans are now putting their foot down on the number of irrelevant ads they are willing to see before they ignore a company completely.” Until now, advertisers utilized users’ web site history and computer cookies to identify which ads to display. All too often these mechanisms represent what happened in the past and usually represent searches or purchases already completed by the consumer.

Image analytics allows for the modeling of future purchases and is predictive not backward looking, like behavioral analytics.


Functional Description of VisualSearch

ImageVision’s VisualSearch Chrome Plug-in uses Flickr Explore image data to demonstrate the power of using color, shades and proportion to perform searches for visually similar image content.

Select An Image

Got to www.Flickr.com/explore
“Right Click” on an image
Select ImageVision VisualSearch

The Visual Search Chrome plug in will take you to www.eyerecognize and automatically place the Image URL in the URL bar, display the image and extract colors.

Palette Tiles, Color Wheel and Color Tiles

The palette tile colors are the base colors for the palettes discovered, not an exact match to any color in the image. The palettes include all the tints, tones and shades for the colors found in the image, and exact matches and proportions are displayed in the color wheel. The tiles are also ordered by proportion of the color in the image.

The first palette tile represents the dominant palette in the image. The first color tile represents the dominant color within that palette.

The color wheel displays all the colors found in a given palette including near black shades, near white tints, and near neutral greys, in the exact proportions they are found relative to that palette within the image. Mouse-over the other palette tiles to display the detail color wheel and color tiles for that palette.

A red frame appears around the first palette indicating that it has been selected as search criteria. The first palette tile and is the default palette selection placed in the color search bar.

Default & Customized VisualSearch

Select whole palettes or individual colors to search by clicking on the palette tile or clicking on individual color tiles above the color wheel. A red frame will highlight your color and palette selections. Palettes and colors, when selected are added to the color search bar in the exact proportion they are found in the image. De-select by again, clicking the appropriate tile, this will remove the frame.

Selecting palettes and colors in the given proportions will return an exact match and matches to visually similar images.

Searches can be modified or customized within the search bar by manipulating the palette and color proportions by dragging the right edge of an individual palette or color. An arrow will appear on the right edge of the palette or color to indicate that it can be dragged right left to change proportion. This changes the color search criteria and will change results from the default settings.

www.ImageVision.com
www.eyerecognize.com