Computer vision

upright=1.35|thumb|This CAPTCHA (GIMPY-R, c. 2005) of "smwm" obscures its message from computer interpretation by twisting the letters and adding a slight background color gradient.

computerized information extraction from images

regularized type of feed-forward neural network that learns features by itself via filter (or kernel) optimization

processing of digitally-represented images with algorithms

detecting visual inputs computationally

mapping of images into a coherent coordinate system

thumb|Some eigenfaces from AT&T Labs|AT&T Laboratories Cambridge

extraction of information from images via digital image processing techniques

U-Net is a convolutional neural network that was developed for image segmentation. The network is based on a fully convolutional neural network whose architecture was modified and extended to work with fewer training images and to yield more precise segmentation. Segmentation of a 512 × 512 image takes less than a second on a modern (2015) GPU using the U-Net architecture.

machine learning algorithm for vision processing

analyzing a real-world object or environment to collect data on its shape and possibly its appearance (e.g. colour)

weighted average/moment of some pixel intensities

mobile computer vision platform for Android developed by Google

digital image capture and processing techniques that use digital computation instead of optical processes

binocular cue to determine depth or distance of an object

problem setup in machine learning, where at test time, a learner observes samples from classes that were not observed during training, and needs to predict the class they belong to

Machine learning algorithm

state or quality by which an item stands out from its neighbors

artificial intelligence company

INDECT is a research project in the area of intelligent security systems performed by several European universities since 2009 and funded by the European Union. The purpose of the project is to involve European scientists and researchers in the development of solutions to and tools for automatic threat detection through e.g. processing of CCTV camera data streams, standardization of video sequence quality for user applications, threat detection in computer networks as well as data and privacy protection.

sensor that can create 3D scans using a patten of light

type of multi-scale signal representation

AI neural network training supercomputer

representation of the distribution of colors in an image

3D reconstruction technique using machine-learning

reconstruction

algorithm used to determine the modes of distribution of probability of a dataset

Framework for multi-scale signal representation

computer vision algorithm

statistical model of an object's shape

portable, artificial vision device

process of combining relevant information

Computer vision framework

right|300px|thumb|Some superquadrics. In mathematics, the superquadrics or super-quadrics (also superquadratics) are a family of geometric shapes defined by formulas that resemble those of ellipsoids and other quadrics, except that the squaring operations are replaced by arbitrary powers. They can be seen as the three-dimensional relatives of the superellipses. The term may refer to the solid object or to its surface, depending on the context. The equations below specify the surface; the solid is specified by replacing the equality signs by less-than-or-equal signs.

right|400px|thumb|Superellipsoid collection with exponent parameters, created using POV-Ray. Here, e = 2/r, and n = 2/t (equivalently, r = 2/e and t = 2/n).