Tutorials#

These tutorials demonstrate how to use Intel® Edge Data Collection.

Tutorial 1: Human annotation #

In this tutorial, you will learn how to perform human annotation with Intel® Edge Data Collection. You will change the configuration in order to select the required images for annotation that would be performed with the tool CVAT (Computer Vision Annotation Tool)

Time to Complete#

15 minutes

Learning Objectives#

  • By the end of this tutorial, you will be able to perform human annotation with Intel® Edge Data Collection.

Prerequisites#

  • Expected prior knowledge in CVAT (Computer Vision Annotation Tool)

  • Complete all steps in the GetStarted section.

Steps#

  • Open the Configuration Management Etcd UI by going to https://<host_IP>:7071/etcdkeeper and go to the section /DataCollectionMicroservice/config. For logging in, the username would be root and password would be what you specified for etcdroot_password in GetStarted section.

  • Specify a value in seconds for the field last_x_seconds_of_images here in order to retrieve the last “x” seconds worth of images for creating human annotation tasks in CVAT

  • Provide an appropriate value for auto_sync_interval_in_sec here that determines how often CVAT needs to be polled for gathering changed or newly created annotations in order to update them in the Data Store

  • In order to understand how to try out more configuration settings, click here

  • Click on Save content on the top right corner.

  • Open CVAT by going to https://<IP>:8443. Replace <IP> with CVAT_HOST that you provided in the file intel_edge_data_collection/Intel_Edge_Data_Collection_1.2.0/IEdgeInsights/build/.env when following the GetStarted section. Login to CVAT using the username and password mentioned in the same file.

  • In CVAT, for the task(s) that are created, review and make changes to corresponding annotations. Click on save. CVAT UI view

  • The performed annotations get saved back into the db in the DataStore microservice.

Summary#

In this tutorial, you learned how to perform human annotation with Intel® Edge Data Collection.

Tutorial 2: Remote data backup #

In this tutorial, you will learn how to perform remote data backup with Intel® Edge Data Collection. You can further use the backed up data as is for various purposes as per your needs.

Time to Complete#

15 minutes

Learning Objectives#

  • By the end of this tutorial, you will be able to perform remote data backup with Intel® Edge Data Collection

Prerequisites#

Steps#

  • Open the Configuration Management Etcd UI by going to https://<host_IP>:7071/etcdkeeper and go to the section /DataCollectionMicroservice/config. For logging in, the username would be root and password would be what you specified for etcdroot_password in GetStarted section.

  • Set mode to 2 as detailed here to activate remote data backup. This would trigger data backup from the Data Store microservice to the user’s workspace.

  • Provide an appropriate value for auto_sync_interval_in_sec here that determines how often remote data backup would be triggered.

  • Click on Save content on the top right corner.

  • Once, remote backup starts happening, the backed-up up images would be found under /opt/intel/eii/data/remote_storage/<topic-name-configured-in-remote-data-backup>/<day-wise-folder-as-YYYYMMDD>/image_dir and their corresponding annotations would be found in the json file present inside /opt/intel/eii/data/<topic-name-configured-in-remote-data-backup>/<day-wise-folder-as-YYYYMMDD>/annotations

    • Example: The backed up images would be found inside /opt/intel/eii/data/edge_video_analytics_results/20230724/image_dir and the corresponding annotation file would be found inside DCAAS/remote-data-backup/edge_video_analytics_results/20230724/annotations Remote data backup view

  • In order to try out more configuration settings such as enabling snapshot mode of remote data backup, enabling remote data backup of object detection data, enabling remote data backup of classification data etc., click here

Summary#

In this tutorial, you learned how to perform remote data backup with Intel® Edge Data Collection.

Learn More#

  • Understand the components, services, architecture, and data flow, in the Overview.

  • Follow instructions for getting started with the microservice in the Get Started Guide.

Known Issues#

  • Sync of manually reviewed annotations and remote backup of annotations/images is dependent on availability of that particular data in DataStore as per its retention policy.

  • Intel® Edge Data Collection waits 5 minutes for DataStore to provide a response for queried data. After the elapsed time, Intel® Edge Data Collection doesn’t error out in the hope that during the next iteration of the query API call to DataStore, meaningful results will be obtained.

  • Intel® Edge Data Collection does not do color conversion of the input frame i.e., BGR to RGB or vice versa. Any desired color format for processing in the UDF that is plugged in to Intel® Edge Data Collection needs to be taken care of in the UDF itself. All images are considered to be in BGR mode by default in Intel® Edge Data Collection.