You have reached the homepage of Derek T. Anderson. I am an Associate Professor at the University of Missouri-Columbia in the Electrical Engineering and Computer Science (EECS) department. On this site, you will find information such as my CV, publications, students, research grants, and laboratory. If you have any additional questions, contact me at Email.

Graduate and undergraduate opportunities: If you are interested in joining my Mizzou INformation and Data FUsion Laboratory (MINDFUL), press the following button to find more details.

Fuzz-IEEE 2019 in New Orleans

On behalf of the Fuzz-IEEE 2017 Organizing Committee, it is our greatest pleasure to invite you to the 2019 IEEE Conference on Fuzzy Systems which will be held in the magnificent city of New Orleans, Louisiana, USA.

More information can be found at here.


A Forensic Anthropology User Interface for Automating Search using Remotely Sensed Data from Unmanned Aerial Vehicles: Preliminary Findings

Description: At the American Association of Physical Anthropologists (AAPA) conference in Austin, TX (April 11, 2018 to April 14, 2018), we will present our unique approach and findings to date for automating detection and documentation of clandestine graves and surface remains (CGSR) using unmanned aerial vehicles and multiple sensors (hyperspectral, thermal and structure from motion). More details will be posted here soon.


New Article (IEEE): Data-Driven Compression and Efficient Learning of the Choquet Integral (link)

Abstract: The Choquet integral (ChI) is a parametric nonlinear aggregation function defined with respect to the fuzzy measure (FM). To date, application of the ChI has sadly been restricted to problems with relatively few numbers of inputs; primarily as the FM has 2^N variables for N inputs and N(2^(N-1)-1) monotonicity constraints. In return, the community has turned to density-based imputation (e.g, Sugeno lambda-FM) or the number of interactions (FM variables) are restricted (e.g., k -additivity). Herein, we propose a new scalable data-driven way to represent and learn the ChI, making learning computationally manageable for larger N. First, data supported variables are identified and used in optimization. Identification of these variables also allows us recognize future ill posed fusion scenarios; ChIs involving variable subsets not supported by data. Second, we outline an imputation function framework to address data unsupported variables. Third, we present a lossless way to compress redundant variables and associated monotonicity constraints. Last, we outline a lossy approximation method to further compress the ChI (if/when desired). Computational complexity analysis and experiments conducted on synthetic data sets with known FMs demonstrate the effectiveness and efficiency of the proposed theory.


New Article (Book Chapter): Fuzzy Choquet Integration of Deep Convolutional Neural Networks for Remote Sensing

Abstract: What deep learning lacks at the moment is the heterogeneous and dynamic capabilities of the human system. In part, this is because a single architecture is not currently capable of the level of modeling and representation of the complex human system. Therefore, a heterogeneous set of pathways from sensory stimulus to cognitive function needs to be developed in a richer computational model. Herein, we explore the learning of multiple pathways--as different deep neural network architectures--coupled with appropriate data/information fusion. Specifically, we explore the advantage of data-driven optimization of fusing different deep nets--GoogleNet, CaffeNet and ResNet--at a per class (neuron) or shared weight (single data fusion across classes) fashion. In addition, we explore indices to open up the data fusion solution and see the importance of each network, how they interact and what aggregation they learned. Experiments are provided in the context of remote sensing for the UC Merced and WHU-RS19 benchmark data sets. In particular, we show that their fusion is the top performer, each network is needed across the various target classes, and unique aggregations (i.e., not common operators) are learned.


Open Source Fuzzy Integral Library

For our Fuzz-IEEE 2017 tutorial "Fuzzy Set Theory in Computer Vision" (link) we created an open source fuzzy integral and computer vision library (link) in Octave and Matlab. The codes at link have also been used to fuse a set of heterogeneous architecture deep convolutional neural networks (DCNNs) for object detection and land classification in remote sensing (see the publications page). Codes and web tutorials are given for the integral, its data-driven learning, visualization; plotting routines, Shapley index, indices of introspection, etc. Computer vision and machine learning examples are also given for hand crafted features and machine learned features using MatConvNet and TensorFlow. The web resources exist to educate. Enjoy!


Fuzz-IEEE Plenary Slides

My slides for the talk "Fusion here, there and almost EVERYWHERE in computer vision - driving new advances in fuzzy integrals" can be found at (link)