Skip to content

Latest commit

 

History

29 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Project Dependency Installation

Install the dependencies with the following command:

pip install -r requirements.txt

Preliminary Exploration Experiments

Comparison experiment command (batch-estimates the performance of each module on CPU and GPU in the early stage, demonstrating the GPU acceleration effect)

Run the following command for the comparison experiments:

python main.py benchmark images --stagewise --with_match

Static Module Partitioning Experiment

Visualize the detection results of three images in one go (under the baseline setting, pick the optimal scheme among the four partitioning options):

python rpc\test_examples_win.py test1.jpg --images test2.jpg test3.jpg

Dynamic Module Partitioning Experiment

Calibrate the model parameters following the instructions in CALIBRATE_README.md.

Dynamically select the optimal scheme based on network latency and memory usage:

python simulate.py images\test1.jpg --find-optimal --process-usage 0 --network-delay 0

Verify whether the scheme chosen by the system above is correct:

python simulate.py images\test1.jpg --test-all --process-usage 0 --network-delay 0

About

A gRPC-based distributed face recognition system that enables dynamic module partitioning by factoring in CPU utilization and network latency.

Topics

Resources

Stars

6 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages