Extracting visual data. Making images Searchable!

The biggest storage challenge for images is that you cannot tag and search them automatically. AIspace automates this process!

This can only be done using AI and is especially useful if your work is media intensive!

Current platforms that allows image search requires you to tag all your images manually! AIspace changes that by applying Artificial Intelligence to automate this process. Along with our Search Engine, we will index your images with tags created automatically and make them searchable! AIspace’s Image AI works in 2 ways:

  • Computer Vision - recognizing thousands of objects in 86 categories.

  • Optical Character Recognition (OCR) - recognizing text in nearly 200 languages (Available soon!)

Computer Vision

AIspace’s computer vision engine analyzes images and return results based on thousands of recognizable objects, living beings, scenery, and actions. In cases where tags may be ambiguous or not common knowledge, our system provides “hints” to clarify the meaning of the tag in context of a known setting. More importantly, it provides attempts to provide context to what the image is about. 

Our computer vision automatically generate the following categories of image data:

  • Description - Generate a description of an entire image in human-readable language, using complete sentences.

  • Visual features - Identify and tag visual features from a set of thousands of recognizable objects, living things, scenery, actions, etc. It also includes the setting (indoor or outdoor), furniture, tools, plants, animals, accessories, gadgets, and so on.

  • Objects - If detects objects like computer, cup, dog, cat and person from 86 object categories. Identify and categorize an entire image, using a category taxonomy with parent/child hereditary hierarchies.

  • Faces - Detect faces in an image and provide information like gender and age.

  • Famous people - Use a domain model to determine if the people detected in the image are known celebrity or famous person.

  • Image types - Uses domain models to detect and identify domain-specific content in an image, such as celebrities, landmarks, clip art, etc.

  • Colour scheme - Analyze color usage and identify the dominant and accent colors.

The following diagram shows the 86 categories that our Image AI can identify and tag. 

babbobox-search-image-categories.jpg

Cloud OCR (Optical Character Recognition)

AIspace converts images to searchable PDF. Based on industry leading OCR technology, AIspace supports the recognition of printed text in nearly 200 languages, including most Asian languages: Chinese, Japanese, Korean, Malay, Vietnamese, Thai and others.

Advanced PDF conversion features reduce the size of scanned images up to 10 times, while preserving the visual quality via powerful MRC compression technology and support of searchable PDF and PDF/A formats.

Note: Supported image formats: JPEG, PNG, GIF, BMP, TIFF.

Handwriting Recognition (Coming Soon!)

Many consider handwriting recognition to be the holy grail of character recognition technology. Watch this space while we work towards bringing you this piece of ground-breaking technology!

AIspace is dedicated to bringing for Image AI to meet your future business need!