Model-free Fitting of Human Stimulus-Response Data

Software packages for non-parametric fitting of a psychometric function, extracting thresholds and slopes, and estimating standard errors are available here soon for both MATLAB and R environments.

Further information could be found on Prof David Foster personal pages

Contact: Prof David Foster
Tel: +44 (0) 161 306 4788

Visualisation Induced SOM (ViSOM)

When used for visualisation of high dimensional data, the self-organising map (SOM) requires a colouring scheme such as U-matrix to mark the distances between neurons. Even so, the structures of the data clusters may not be apparent and their shapes are often distorted. In this paper, a visualisation- induced SOM (ViSOM) was proposed in 2000 as a new tool for data visualisation. The algorithm constrains the lateral contraction forces between a winning neuron and its neighbouring ones and hence regularises the inter-neuron distances. The mapping preserves directly the inter-neuron distances locally on the map along with the topology. It produces a graded mesh in the data space and can accommodate both training data and new arrivals. The ViSOM represents a class of discrete principal curves and surfaces.

Proposed in June 2000.

Original IEEE TNN paper to view in PDF format "A Novel Method for Multivariate Data Projection and Structure".

Software (Matlab code) download - available soon.

Contact: Dr Hujun Yin
Tel: +44 (0) 161 306 8714
Hujun Yin's personal web page

Binary Gradient Correlation Patterns MATLAB package

Matlab Code for Implementing BGCP based methods (BGCP and BGCPM) for extract binary features robust face recognition, as well as demo code of recognition on various databases such as:

YaleB:  BGCP with single gallery samples; BGCPM with single gallery sample
AR:  BGCP and BGCPM with single gallery sample;  BGCP and BGCPM with multiple gallery samples
FERET:  BGCPMLDA and BGCPM with FDA learning processing
LFW:  BGCP and BGCPM for face verification

Software (Matlab code) downloadable - upon request.

Contact: Hujun Yin
Tel: +44 (0) 161 306 8714
Hujun Yin's personal web page

▲ Up to the top