Vision and color science

AfterEffects is a web-based platform containing code and data for generating motion, color, and texture after-effects. It reproduces these effects according to infomax or error minimization organization goals using Sequential Principal Curves Analysis (SPCA).

COLORLAB is a color computation and visualization toolbox to be used in the MATLAB environment. It is intended for colorimetric applications, such as color image processing and psychophysical experimentation. The toolbox includes classical Tristimulus representations (e.g. CIE XYZ 1931), current Color Appearance Models (CIE Lab 76 or CIECAM2000), and various tools for visualization and color processing.

CSFTools is a MATLAB toolbox that provides spatial Contrast Sensitivity Functions (CSF) of the ModelFest Standard Spatial Observer (Watson&Malo 2002). It also includes chromatic CSFs from Mullen (1985). The toolbox covers spatio-temporal CSF in terms of frequency (Kelly 1979) and speed (Daly 2000).

DNTools is a MATLAB toolbox that models texture perception using Orthonormal Wavelets and Divisive Normalization. It simulates a V1 cortex model tuned to image quality opinions and is found to perform non-linear ICA on natural images.

LinearNoisyTools is a MATLAB toolbox for modeling texture perception, focusing on Poisson noise in V1 sensors. The toolbox includes achromatic CSF filtering, steerable wavelets with optimized architectures, and psychophysically measured noise models.

The Human Camera is a Brain-Machine Interface designed for image transmission. It uses the human brain for feature extraction and dimensionality reduction, instead of conventional algorithms. The toolbox showcases decoding algorithms based on a V1 cortex model.

Topographic Independent Component Analysis (TICA) is a method for learning topographically organized features from visual data. It extends the Independent Component Analysis (ICA) algorithm by incorporating spatial proximity, aiming to model how neurons in the visual cortex represent information with localized and continuous receptive fields. TICA has been applied to understand orientation domains in the brain’s visual cortex, but studies have shown that the model may produce discontinuous, scrambled orientation maps, contradicting experimental observations.

VirtualNeuroLabs provides a collection of MATLAB tools designed to simulate physiological and psychophysical experiments. It also illustrates the behavior of neural models when processing complex stimuli. These tools are ideal for students studying visual neuroscience at a graduate level.