Purdue Research Foundation

Charter Communications Operating, LLC

Yu Charlie Hu

Abhinav Pathak

Ning Ding

Samuel Midkiff

Pranab Dash

Abhilash Jindal

Zhaoning Jonny Kong

Dhananjay Lal

Dell Wolfensparger

Mun G Choi

Christopher G Turner

Viviane Espinoza McLandrich

Jiayi Meng

An offloading schedule for offloading image frames that are to be generated by an AR device in a subsequent time period is determined, the offloading schedule identifying certain image frames generated in the subsequent time period that are to be offloaded and certain image frames generated in the subsequent time period that are not to be offloaded, the offloading schedule being selected from a plurality of offloading schedules based on a tracking stride of the offloading schedule. The AR device sends to a computing device at least some of the image frames generated in the subsequent time period in accordance with the offloading schedule.

Optimizing image offloading in edge-assisted augmented reality

An augmented reality (AR) device captures a first image frame of a first scene in a real world. The AR device compresses the first image frame using a same bitrate across an entirety of the first image frame to generate a first compressed image frame. The AR device wirelessly sends, to a server, the first compressed image frame. The AR device receives, from the server, object detection results that identify locations of objects depicted in the first image frame.

Efficient offloading of video frame processing tasks in edge-assisted augmented reality

A method for generating an energy usage profile for a software program executed in a computing device includes generating a call trace of an executed system call, identifying a first power consumption and duration of a first power state due to the first system call using a model, identifying a second power consumption and duration of a second power state based on the model, and generating an energy usage profile for the software program. The energy usage profile includes energy consumption of the computing device for the system call based on the first power consumption level multiplied by the first duration and the second power consumption level multiplied by the second duration associated an identifier for the call trace.

System and method for energy usage accounting in software applications

A method of determining a model for pixel power consumption for each pixel in a display of a device displaying each color in a color space is disclosed which includes establishing a color space for the display, decomposing the color space into a plurality of subgrids, measuring the pixel power associated with a selected set of colors in each subgrid of the plurality of subgrids, establishing a pixel power model for each subgrid of the plurality of subgrids by applying a function to the power values at the selected set of colors in that subgrid, and deriving a piecewise pixel power model for the entire color space which includes pixel power models for the plurality of subgrids.

Method and system of profiling display power

A method for profiling energy usage in graphics user interfaces (UI) in handheld mobile devices is disclosed, which includes quantifying the central processing unit (CPU) energy drain of each UI update, quantifying the graphics processing unit (GPU) energy drain of each UI update, quantifying the number of pixels changed due to each UI update, identifying an UI update that consumes energy drain but results in no pixel changes to the displayed frame as a graphics energy bug.

Graphics energy profiling system and method

A method for profiling energy usage of invoking an application programming interface (API) by an application in a computing device. The method includes obtaining source code for the API, modifying each class by adding a callback function identifier field that is initiated to a unique value upon instantiation of each object that belongs to the class, identifying each location in the source code that posts the callback function for asynchronous execution by enqueueing each object instantiated from the class containing the callback function into the system callback queue, modifying the source code to log the callback function identifier of object at the location that dequeues objects from the system callback queue, modifying source code by adding two system logging function calls to log the callback function identifier of the dequeued object before and after executing the callback function, executing the application, and performing energy accounting of the asynchronous API calls.

System and method for energy profiling asynchronous API calls

System and method for profiling energy usage of API calls that asynchronously invoke functions

Embodiments of the present invention provide a system and methods for detecting power bugs. In one embodiment, a computer-implemented method for analyzing a computer code includes generating a control flow graph for at least a portion of the computer code at a processor. The method further includes identifying power bugs by traversing the control flow graph if the control flow graph exits without performing a function call to deactivate power to any component of a device configured to execute computer executable instructions based on the computer code after performing a function call to activate power.

Systems and methods of detecting power bugs

A method for profiling energy usage in graphics user interface (UI) update in handheld mobile devices is disclosed. The method includes mapping the pixel changes of each frame to a responsible UI update, identifying method invocations at all layers of the UI rendering stack, tracking asynchronous dependence across all layers of the UI rendering stack corresponding to each UI update by assigning a unique frame identification (ID) per frame refresh to track asynchronous dependence between UI updates and corresponding callbacks across all layers of the UI rendering stack, tracking information flow across a corresponding graphics processing unit (GPU) by using an OpenGL record-and-replay method, reconstructing the frames by replaying the OpenGL calls on the GPU during post processing, comparing frames to account for pixel changes, and tracking and attributing the energy usage during UI operations to individual UI updates.

Embodiments of the present invention provide a system and methods for detecting power bugs. In one embodiment, a computer-implemented method for analyzing a computer code includes generating a control flow graph for at least a portion of the computer code at a processor. The method further includes identifying power bugs by traversing the control flow graph if the control flow graph exits without performing a function call to deactivate power to any component of a device configured to execute computer executable instructions based on the compute code after performing a function call to activate power.

Growing community of inventors

West Lafayette, IN, United States of America

Yu Charlie Hu