Perceptual Interfaces are concerned with extending human computer interaction to use all modalities of human perception. Our current research efforts are focused at including vision, audition, and touch in the process. The goal of perceptual reality is to create virtual and augmented versions of the world, that are perceptually identical to the human with the real world. The goal of creating perceptual user interfaces is to allow humans to have natural means of interacting with computers, appliances and devices using voice, sounds, gestures, and touch. In both creating virtual reality, and in acquiring multimodal input from humans, our research emphasizes physics-based algorithms, efficient computation, and real-time implementations.
Another portion of our research is concerned with creating prosthetic devices for the vision and hearing impaired, by mapping inputs from one modality into equivalent ones in another, so that computationally augmented input streams can be created with extra content from the missing modality.
Ph.D., Johns Hopkins University, 1991
B.Tech., IIT Bombay, 1985
Audio and Computational Acoustics:
Acoustics for perceptual reality: Head Related Transfer Functions, Room Impulse Responses, Auditory Telepresence, Reproduction of audio using headphones, Reproduction using speakers. Keynote talk at Ambisonics 2010
Microphone Arrays: Beamforming, Source Localization, Source Modeling, Spherical Microphone Arrays, Cameras and Arrays, The Audio Camera, Non-blind algorithms
Speech: Speaker ID, Large Scale Computing for Speech Processing
Auditory User Interfaces: Sonification of Data, Systems for presentation, Audio in Games
Underwater acoustics: Bubble counting, sound propagation in bubbly media and fog
Scientific and Statistical Computing:
Fast Multipole Methods: Data Structures; Adaptive Algorithms; FMM for the Laplace, Helmholtz, Biharmonic, Maxwell and Stokes kernels; General kernels; Scattering problems
High Performance Computing: GPU and Heterogeneous Parallel Computing
Computational Statistics and Learning Methods: Improved Fast Gauss Transform in high dimensions; Mean-Shift; Particle Filters; "Fast N-Body Learning"; Classification, Ranking, Gaussian Processes, Speaker ID.
Data Fitting and Modeling: RBF interpolation; Data Structures for higher dimensional data; Non Uniform Fast Fourier Transforms, Data Assimilation; Gaussian Process Regression
Boundary Element Methods: Speedup via the FMM; Computation of singular and near singular integrals; Meshless Methods; Software.
Small Angle Scattering: Small Angle X-ray and Neutron scattering, fast algorithms
Other (older stuff): Electrical Impedance Tomography; Bubble Dynamics; Free surface flow; Spectral Methods; Effective Media; Inverse problems.
Vision aware audio; Tracking; Pose; Kernel methods, Vision based prosthetics for the visually impaired, application of fast algorithms to computer vision and image processing, The Audio Camera.