Computer Science & AI

AI for robotics, automated planning, game theory, speech recognition, data visualization, rule-based expert systems and logic programs.

ISR artificial intelligence research on hierarchical task network planning has influenced nearly all subsequent work in this area. We have deep understanding of artificial intelligence planning and the use of mean-field game theory to predict decisions. Our pioneering history in data visualization produced Spotfire, a starfield multidimensional data visualization display tool using dynamic queries that is used for drug discovery, genomic data analysis, oil and gas discovery, manufacturing control, marketing, supply chain management, and financial analysis. We also developed Treemaps, a space-filling method of visualizing large hierarchical collections of quantitative data that gives users the ability to see thousands of data bits in a fixed space that facilitates discovery of patterns, clusters and outliers.  We have expertise in formal methods for description and analysis of concurrent and distributed systems, model checking and abstract interpretation for embedded control and systems biology. Today we are introducing innovations in computer vision and hyperdimensional computing theory for robots, as well as geometric and scientific algorithms for autonomous vehicles, computer graphics, and virtual reality.

Recent publications by ISR computer science and artificial intelligence faculty

2019

GraphRQI: Classifying Driver Behaviors Using Graph Spectrums

Rohan Chandra, Uttaran Bhattacharya, Trisha Mittal, Xiaoyu Li, Aniket Bera, Dinesh Manocha

The GraphRQI algorithm identifies driver behaviors from road agent trajectories. It is 25 percent more accurate over prior behavior classification algorithms for autonomous vehicles.  | Watch a video about GraphRQI |

arxiv.org

Realtime Simulation of Thin-Shell Deformable Materials using CNN-Based Mesh Embedding

Qingyang Tan, Zherong Pan, Lin Gao, and Dinesh Manocha

A new method bridges the gap between mesh embedding and physical simulation for efficient dynamic models of clothes. The key technique is a graph-based convolutional neural network (CNN) defined on meshes with arbitrary topologies, and a new mesh embedding approach based on physics-inspired loss term. After training, the learned simulator runs10–100 times faster and the accuracy is high enough for robot manipulation tasks. | Watch a video about this method |

arxiv.org

 

ISR computer science and artificial intelligence news


Top