fxgsell / gg-edge-inference Goto Github PK

View Code? Open in Web Editor NEW

17.0 17.0 7.0 106.29 MB

Using AWS Greengrass with the Nvidia Jetson TX2 to run ML models prepared with Amazon SageMaker.

License: MIT License

Makefile 1.36% Python 58.90% Jupyter Notebook 39.74%

gg-edge-inference's People

Contributors

Stargazers

Watchers

Forkers

empireofkings wenming amrragab8080 comdaze lawrencechen0921 liuweixinning ruparee

gg-edge-inference's Issues

1-greengrass-deployment: confirming certs installed via stop/start

1-greengrass-deployment

clarifying instructions for installing certs
confirming certs have been properly installed on the device (possibly with device stop/start and showing where files were stored)

2-face-detection - make failure

make execution fails:

ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ make

echo "Zipping..."
Zipping...
rm -f package.zip
zip -rq package.zip greengrass_common  greengrass_ipc_python_sdk  greengrasssdk  *.py
echo "Uploading to Lambda"
Uploading to Lambda
aws --profile default --region us-east-1 lambda update-function-code --function-name  --zip-file fileb://`pwd`/package.zip
usage: aws [options] <command> <subcommand> [<subcommand> ...] [parameters]
To see help text, you can run:

  aws help
  aws <command> help
  aws <command> <subcommand> help
aws: error: argument --function-name: expected one argument
make: *** [deploy] Error 2

properties.mk

# Static
PROFILE=default
LAMBDA_ALIAS=latest
AWS_COMMAND=aws --profile $(PROFILE) --region $(REGION)

# Edit these values (where required)
REGION=us-east-1
GG_GROUP=ml-edge-workshop
BUCKET=ml-edge-workshop-shirkeys

Is function_name a value we need to pass? Assuming that could be changed during Lab 1, including it would make sense

2-face-detection - Deployment failing within make

Deployment failing
Deployment failing silently

make output

ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ make
echo "Zipping..."
Zipping...
rm -f package.zip
zip -rq package.zip greengrass_common  greengrass_ipc_python_sdk  greengrasssdk  *.py
echo "Uploading to Lambda"
Uploading to Lambda
aws --profile default --region us-east-1 lambda update-function-code --function-name ml-edge-workshop-lab-1 --zip-file fileb://`pwd`/package.zip
{
    "TracingConfig": {
        "Mode": "PassThrough"
    }, 
    "CodeSha256": "x4otKs1etleHIqP+H5tMfeRe9D5DUhGK44wQ4svy1tk=", 
    "FunctionName": "ml-edge-workshop-lab-1", 
    "CodeSize": 21229, 
    "MemorySize": 128, 
    "FunctionArn": "arn:aws:lambda:us-east-1:306280812807:function:ml-edge-workshop-lab-1", 
    "Version": "$LATEST", 
    "Role": "arn:aws:iam::306280812807:role/ml-edge-workshop-lab-1-role", 
    "Timeout": 3, 
    "LastModified": "2018-06-09T12:14:53.463+0000", 
    "Handler": "lambda_function.lambda_handler", 
    "Runtime": "python2.7", 
    "Description": ""
}
aws --profile default --region us-east-1 lambda publish-version --function-name ml-edge-workshop-lab-1 --query Version --output text  > LAMBDA_VERSION
aws --profile default --region us-east-1 lambda update-alias --name latest --function-name ml-edge-workshop-lab-1 --function-version `cat LAMBDA_VERSION`
{
    "AliasArn": "arn:aws:lambda:us-east-1:306280812807:function:ml-edge-workshop-lab-1:latest", 
    "FunctionVersion": "1", 
    "Name": "latest", 
    "Description": ""
}
rm LAMBDA_VERSION
echo "Deploying to GG"
Deploying to GG
aws --profile default --region us-east-1 greengrass list-groups --query "Groups[?Name=='ml-edge-workshop'].Id" --output text > GROUP_ID
aws --profile default --region us-east-1 greengrass list-groups --query "Groups[?Name=='ml-edge-workshop'].LatestVersion" --output text > GROUP_VERSION
aws --profile default --region us-east-1 greengrass create-deployment --group-id `cat GROUP_ID` --deployment-type NewDeployment --group-version-id `cat GROUP_VERSION` --query "DeploymentId" --output text > GROUP_DEPLOYMENT
aws --profile default --region us-east-1 greengrass get-deployment-status --group-id `cat GROUP_ID` --deployment-id `cat GROUP_DEPLOYMENT` --query "DeploymentStatus" --output text  > STATUS
Build in progress: Building
rm GROUP_ID GROUP_VERSION GROUP_DEPLOYMENT STATUS

manual steps to inspect output of deployment steps (ie. STATUS file before it is deleted)

ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ aws --profile default --region us-east-1 greengrass list-groups --query "Groups[?Name=='ml-edge-workshop'].Id" --output text > GROUP_ID                                 
ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ aws --profile default --region us-east-1 greengrass list-groups --query "Groups[?Name=='ml-edge-workshop'].LatestVersion" --output text > GROUP_VERSION
ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ aws --profile default --region us-east-1 greengrass create-deployment --group-id `cat GROUP_ID` --deployment-type NewDeployment --group-version-id `cat GROUP_VERSION` --query "DeploymentId" --output text > GROUP_DEPLOYMENT
ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ aws --profile default --region us-east-1 greengrass get-deployment-status --group-id `cat GROUP_ID` --deployment-id `cat GROUP_DEPLOYMENT` --query "DeploymentStatus" --output text  > STATUS
ec2-user:~/environment/GG-Edge-Inference/2-face-detection (master) $ cat ./STATUS 
Failure

Executing create-greengrass-config.py from Cloud9 environment is failing to sign the required URLs

ec2-user:~/environment/GG-Edge-Inference/1-greengrass-configuration (master) $ python3 create-greengrass-config.py --create-group ml-edge-workshop --bucket ml-edge-workshop-lab-1 --function ml-edge-workshop-lab-1
Creating IAM role for Greengrass
Traceback (most recent call last):
File "create-greengrass-config.py", line 294, in
state = create_group(args.group_name, args.bucket)
File "create-greengrass-config.py", line 182, in create_group
role, role_policy = create_gg_role(bucket, certificate['certificateArn'][-10:])
File "create-greengrass-config.py", line 125, in create_gg_role
AssumeRolePolicyDocument=json.dumps(assume_role_document)
File "/opt/c9/python3/local/lib/python3.6/dist-packages/botocore/client.py", line 314, in _api_call
return self._make_api_call(operation_name, kwargs)
File "/opt/c9/python3/local/lib/python3.6/dist-packages/botocore/client.py", line 612, in _make_api_call
raise error_class(parsed_response, operation_name)
botocore.exceptions.ClientError: An error occurred (InvalidClientTokenId) when calling the CreateRole operation: The security token included in the request is invalid

via mplayer, not seeing any face detection from 2-face-detection lambda

clarify instructions in Lab 1 for AWS Lambda function creation

Missing attribute from static_config

$ python3 create-greengrass-config.py --create-group ml-edge-workshop --bucket ml-edge-workshop-lab-1 --function ml-edge-workshop-lab-1
Creating IAM role for Greengrass
Generating configuration package
Configuration and certificates generated: certificates.zip
Traceback (most recent call last):
File "create-greengrass-config.py", line 311, in
state = add_function(args.function_name, state)
File "create-greengrass-config.py", line 209, in add_function
function = static_config.FUNCTION_INITIAL_VERSION
AttributeError: module 'static_config' has no attribute 'FUNCTION_INITIAL_VERSION'

Face Recognition Precision

Hi, I wonder how to improve the precision in AWS Rekognition, since I recently tested with multiple people, but sadly returned with same face id.
Sometimes have to wait several minutes until the face was detected as a different one.

Is there any way to fix the code or set similarity threshold higher?
By the way, I am using Raspberry Pi 3 and a logitech c310 webcam.

Thank you!

Error in make for Lab 3 - hybrid

➜  3-hybrid-face-recognition git:(master) ✗ make
make -C function
zip -r hybrid-face-recognition.zip *
  adding: Makefile (deflated 8%)
  adding: lambda_function.py (deflated 61%)
aws cloudformation package \
		--template-file hybrid-face-recognition.yml \
		--s3-bucket my-bucket \
		--output-template-file hybrid-face-recognition-release.yml

Unable to upload artifact ./function/hybrid-face-recognition.zip referenced by CodeUri parameter of FaceRecognition resource.
An error occurred (AllAccessDisabled) when calling the PutObject operation: All access to this object has been disabled
make: *** [build] Error 255

python3 create-greengrass-config.py --ip-address fails

How to use webcam or other cameras instead of onboard camera

Hi, Thanks for the wonderful project.

I want to use the webcam (Logitech C310) instead of onboard camera on jetson tx2.

How to use the webcam video stream from "/dev/video1" for the face detection project?

Likely should split out setup of Cloud9 / local environment as Lab 0 / Setup

greengrass error when stopping/starting on device

When stopping / starting per this page

where:

Restart the AWS Greengrass daemon by running the following commands in your core device terminal.
cd /greengrass/ggc/core/
sudo ./greengrassd stop
sudo ./greengrassd start

then, when executed on device:

nvidia@tegra-ubuntu:/greengrass/ggc/core$ sudo ./greengrassd stop
[sudo] password for nvidia:
nvidia@tegra-ubuntu:/greengrass/ggc/core$ sudo ./greengrassd start
Setting up greengrass daemon
Validating hardlink/softlink protection
Validating execution environment
Found cgroup subsystem: cpuset
Found cgroup subsystem: cpu
Found cgroup subsystem: cpuacct
Found cgroup subsystem: blkio
Found cgroup subsystem: memory
Found cgroup subsystem: devices
Found cgroup subsystem: freezer
Found cgroup subsystem: net_cls
Found cgroup subsystem: perf_event
Found cgroup subsystem: net_prio
Found cgroup subsystem: pids
Found cgroup subsystem: debug

Starting greengrass daemon


Greengrass daemon 11306 failed to start
Failed to parse config.json. err: UseSystemd [[yes|no]] is invalid, please use "yes" or "no"
nvidia@tegra-ubuntu:/greengrass/ggc/core$

and when catting config.json:

nvidia@tegra-ubuntu:/greengrass/ggc/core$ cat /greengrass/config/config.json
{
    "coreThing": {
        "caPath": "[ROOT_CA_PEM_HERE]",
        "certPath": "[CLOUD_PEM_CRT_HERE]",
        "keyPath": "[CLOUD_PEM_KEY_HERE]",
        "thingArn": "[THING_ARN_HERE]",
        "iotHost": "[HOST_PREFIX_HERE].iot.[AWS_REGION_HERE].amazonaws.com",
        "ggHost": "greengrass.iot.[AWS_REGION_HERE].amazonaws.com"
    },
    "runtime": {
        "cgroup": {
            "useSystemd": "[yes|no]"
        }
    },
    "managedRespawn": false
}