Products

Image

The world’s first true Reverse Image Search technology that actual works.

Photon, a revolution reverse image search solution powered by DejaVuAI®, allows rapid searching 1:N, N:1, or M:N, regardless of image adjustments.

Photon Online

Best For: individual users searching standard quality

Photon Online provides standard quality rating searching and is the perfect solution for individual users managing their own portfolio of images. Easily upload your images for daily searching against known sites:

  • Alibaba.com
  • Aliexpress.com
  • Soundcloud.com
  • Etsy.com
  • Redbubble.com
  • Fineartamerica.com
  • Craigslist
  • Printrunner
  • Greatbigcanvas.com
  • Wayfair.com
  • Pintrest.com
  • Hottopic.com
  • Spencersonline.com
  • EthanAllen.com
  • HGTV.com
  • Walmart.com
  • Target.com
  • Macys.com
  • Livingspaces.com
  • Oldnavy.com
Image
Image

Photon On Prem

Best For: searching a library images needing more quality options

Available Options: Personal Perpetual | Personal Subscription | Enterprise Subscription

Photon is less than 90MB and can easily be loaded on low SWAP processing units. The CPU can be anything from an ARM Processor (Raspberry Pi) to cluster computer or even super computers, Photon is algorithm-based. The faster the CPU and the more RAM and cache you throw at it, the faster it can return results. The DejaVuAI® algorithm has an exceptionally small footprint, such that images can be processed for 1:n, n:1, or m:n image matching. Our solution also automatically identifies scale and rotation of the matched images, providing immediate image registration.

Photon’s Low SWAP-capable image matching accuracy is versatile for a wide range of applications, military, commercial, and civilian government such as state and local police

  • Provides fast and accurate decision-quality intelligence for all levels of command.
  • Works in environments with either limited reconnaissance intel or in changed environments due to manmade or natural adjustments/damage.
  • Dissects images and identifies source images, regardless of object, filter, part, foreground, middle-ground, or background.
  • Enables low-cost devices, produced en masse, to be exponentially more accurate with IFF recognition, navigation, and identification.
  • Builds AI trustworthiness through accurate image recognition before it gets to other ML/AI components.
  • Provides accurate and rapid image recognition on low-power (e.g., wearable or UAV) compute that does not require any GPU processing.
  • Supports cross-imager image registration and matching across a variety of imager types including LIDAR, radar, terahertz, IR, and thermal.
  • Highly flexible and scalable for a wide range of deployment types for implementations both board-level CPU-based computers or large cluster/VM environments. We anticipate integration with ARM, M1, and Android processors