Arizona State University

Cornell University

Lawrence Livermore National Security, LLC

Northern Arizona University

Suren Jayasuriya

Andreas Spanias

Alyosha C Molnar

Odrika Iqbal

Sriram Sivaramakrishnan

Visar Berisha

Rushil Anirudh

Hyojin Kim

Kyle Champley

Sameeksha Katoch

Pavan Turaga

Mark Buckler

Adrian Sampson

Gautam Dasarathy

Kadri Aditya Mohan

Jianwei Zhang

John Janiczek

Philip Christensen

Christopher Edwards

Sreenithy Chandran

Divya Mohan

Albert William Reed

Alyosha Molnar

Tracking-based motion deblurring via coded exposure is provided. Fast object tracking is useful for a variety of applications in surveillance, autonomous vehicles, and remote sensing. In particular, there is a need to have these algorithms embedded on specialized hardware, such as field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs), to ensure energy-efficient operation while saving on latency, bandwidth, and memory access/storage. In an exemplary aspect, an object tracker is used to track motion of one or more objects in a scene captured by an image sensor. The object tracker is coupled with coded exposure of the image sensor, which modulates photodiodes in the image sensor with a known exposure function (e.g., based on the object tracking). This allows for motion blur to be encoded in a characteristic manner in image data captured by the image sensor. Then, in post-processing, deblurring is performed using a computational algorithm.

Tracking-based motion deblurring via coded exposure

Various embodiments of a system for linear unmixing of spectral images using a dispersion model are disclosed herein.

Systems and methods for differentiable programming for hyperspectral unmixing

Systems, methods, and apparatuses to restore degraded speech via a modified diffusion model are described. An exemplary system is specially configured to train a diffusion-based vocoder containing an upsampler, based on pairing original speech x and degraded speech mel-spectrum msamples; train a deep convoluted neural network (CNN) upsampler based on a mean absolute error loss to match the estimated original speech {circumflex over (x)}′ outputted by the diffusion-based vocoder by extracting the upsampler, generating a reference conditioner, and generating a weighted altered conditioner c′. The system further optimizes speech quality to invert non-linear transformation and estimate lost data by feeding the degraded mel-spectrum mthrough the CNN upsampler and feeding the degraded mel-spectrum mthrough the diffusion-based vocoder. The system then generates estimated original speech {circumflex over (x)}′ based on the corresponding degraded speech mel-spectrum m. Other related embodiments are described.

Systems, methods, and apparatuses for restoring degraded speech via a modified diffusion model

A system for generating a 4D representation of a scene in motion given a sinogram collected from the scene while in motion. The system generates, based on scene parameters, an initial 3D representation of the scene indicating linear attenuation coefficients (LACs) of voxels of the scene. The system generates, based on motion parameters, a 4D motion field indicating motion of the scene. The system generates, based on the initial 3D representation and the 4D motion field, a 4D representation of the scene that is a sequence of 3D representations having LACs. The system generates a synthesized sinogram of the scene from the generated 4D representation. The system adjusts the scene parameters and the motion parameters based on differences between the collected sinogram and the synthesized sinogram. The processing is repeated until the differences satisfy a termination criterion.

Reconstruction of dynamic scenes based on differences between collected view and synthesized view

Various embodiments of systems and methods for adaptive video subsampling for energy-efficient object detection are disclosed herein.

Adaptive video subsampling for energy efficient object detection

Various embodiments for system and method of adaptive lighting for data-driven non-line-of-sight imaging are disclosed.

Systems and methods for adaptive lighting for data-driven non-line-of-sight three-dimensional capabilities

A depth of field imaging apparatus includes a light field imager and a time of flight imager combined in a single on-chip architecture. This hybrid device enables simultaneous capture of a light field image and a time of flight image of an object scene. Algorithms are described, which enable the simultaneous acquisition of light field images and a time of flight images. Associated hybrid pixel structures, device arrays (hybrid imaging systems), and device applications are disclosed.

Depth field imaging apparatus, methods, and applications

A configurable image processing system can process image data for multiple applications by including an image sensor capable of operating in a machine vision mode and a photography mode in response to an operating system command. When operating in machine vision mode, the image sensor may send image data to first processor for machine vision processing. When operating in photography mode, the image sensor may send image data to an image coprocessor for photography processing.

Configurable image processing system and methods for operating a configurable image processing system for multiple applications

Growing community of inventors

Tempe, AZ, United States of America

Suren Jayasuriya