SSRU69_cd - Documentation Index

Table of Contents

• Welcome
• What's new
  • ... in Maximo Visual Inspection feature channel
  • What's new in Maximo Visual Inspection in 9.2
  • … in 9.1
  • … in 9.0
  • … in 8.9
  • … in 8.8
  • … in 8.7
  • … in 8.6
  • ... in fix packs
• Overview
  • User interface
    • Logging in
    • Determining the product version
  • Use cases
  • REST APIs
  • Custom resource status codes
  • Models and supported functions
  • Users and roles
  • Concepts
  • Documentation conventions
  • Accessibility
  • Scenarios
    • Scenario: Detecting objects in images
    • Scenario: Detecting objects in videos
    • Scenario: Classifying images
    • Scenario: Detecting segmented objects in images
    • Scenario: Detecting objects in high-resolution images
    • Scenario: Detecting anomalous objects in images
    • Scenario: Detecting actions in videos
    • Scenario: Adding preprocessing and postprocessing
  • Deploying on Azure
• Getting started
  • Creating a data set
  • Labeling objects
    • Creating labels
    • Labeling an image
    • Labeling a video
    • Labeling an action
    • Classifying images
  • Training a model
  • Deploying a trained model
• Planning
  • GPUs
    • GPU requirements
    • GPU storage
    • Supported GPU devices
    • GPU splitting with NVIDIA Multi-Instance GPU
    • Deployment of one model onto multiple GPUs
    • CUDA compute capability
  • Storage requirements
  • COCO and VOC annotations
  • Image size and resolution
  • Supported web browsers
• Upgrading
• Installing
• Configuring
  • Data sets
    • Data set types
      • Visual prompting
      • Training
      • Inspection
    • Data sets for object detection models
    • Data sets for anomaly optimized models
    • Images and videos
      • Supported image files
      • Supported video files
      • Image resolution for model training and inference
      • Image metadata
      • Importing images with COCO annotations
      • Searching for images and videos in data sets
      • Moving or copying images and videos
  • Labeling objects
    • Number of required images
    • Editing label and category names
    • Label, resource, and class names
    • Identifying and drawing objects
    • Labeling actions in videos
    • Editing boxes and polygons
  • Visual prompting models
    • Enabling the visual prompting service
    • Generating a visual prompting model
    • Visual prompting model workflows
  • Training models
    • Model types
      • Image classification
      • Object detection
        • Faster R-CNN
        • YOLOv3
        • Tiny YOLOv3
        • Detectron2
        • High-resolution
        • Single Shot Detector
        • Anomaly optimized
      • Action detection
      • Advanced settings
    • Graph for training models
    • Included base models
    • Custom models
      • Preparing custom models to train data sets
        • class MyTrain(TrainCallback)
        • Object detection callbacks
        • Classification callbacks
      • Preparing custom models for deployment
        • class MyDeploy(DeployCallback)
  • Deploying trained models
    • TensorRT models
      • TensorRT model conversion
  • Testing models
  • Model validation
  • Refining models
    • Auto labeling objects
      • Auto labeling objects in data sets
      • Auto labeling videos
    • Augmenting data sets
      • Augmentation settings
  • Production work flow
    • Automatically deploying the newest model
  • Project groups
  • Preprocessing and postprocessing
  • Importing, exporting, and downloading information
  • REST APIs
    • Generating an API key
    • API key tokens
  • Understanding metrics
• Managing
  • Sharing resources
  • Data retention policy
  • Workload scale customization
  • Object storage
    • Enabling object storage
    • Object storage limitations
    • Debugging object storage issues
  • Edge manager
    • Single sign-on
    • Running diagnostics
• Administering
  • Managing SSL certificates
  • Backing up data
• Troubleshooting
  • Data set issues
    • Auto-labeling generates an error message
    • Uploading a large file fails
    • Action detection issues
  • Training and deployment issues
    • Model training is not working
    • Model training fails and generates an error message
    • Model training and inferencing fail
    • Model accuracy value is unexpected
    • Model import generates an error message
    • Training or deployment does not complete
    • Model deployment fails
    • Deployed models are stuck in starting state
  • GPU issues
    • Resolving GPU availability issues
    • Nodes have no available GPUs
    • Cannot create a driver container
    • Access to fewer GPUs
  • User interface issues
    • User interface does not work
    • Unexpected or old pages are displayed
    • Cannot upload or play a video
    • Forgotten username or password
    • API keys do not work
    • Cannot import DICOM format file
    • No data is showing when connected to IBM Watson IoT Platform
    • Resource pages are not populated
    • Postgres Kubernetes pod fails to start
    • Thumbnails do not load or images previews are missing
    • Troubleshooting MQTT broker connection errors
  • Frequently asked questions about graphics processing units
  • Gathering logs and contacting support
    • Audit log samples
  • Contacting IBM Support
• Maximo Visual Inspection Edge
  • Overview
    • What's new
      • ... in Maximo Visual Inspection Edge feature channel
      • … in 9.2
      • … in 9.1
      • … in 9.0
      • … in 8.9
      • … in 8.8
      • … in 8.7
      • … in 8.6
      • ... in fix packs
    • Concepts
    • Interaction with Maximo Visual Inspection
    • REST APIs
    • Dashboard view
    • Scenarios
      • Scenario: Creating an inspection to collect images for model training
      • Scenario: Creating an inspection to inspect images and display inspection results
      • Scenario: Downloading, deploying, and using a new model
      • Scenario: Performing a model inference
  • Getting started
    • Getting started for users
      • Adding an API key
      • Adding an input source
      • Creating a station
      • Configuring an inspection
    • Getting started for administrators
      • Connecting with Maximo Visual Inspection
      • Adding users
      • Connecting with MQTT
      • Connecting with Twilio
  • Planning
    • Platform requirements
    • Installing NVIDIA Container Toolkit and Docker
  • Installing, uninstalling, and seamlessly upgrading
    • Installing
    • Uninstalling
  • Internal MQTT broker
    • MQTT messaging in internal processes, triggers, and alerts
    • Integrating external systems with the internal MQTT broker
    • Using TLS to encrypt outgoing alert messages
    • MQTT limitations
  • Configuring
    • Configuration components
      • Input sources
        • Adding a GigE Vision camera
      • Stations
      • Models
        • Anomaly detection models
        • Composite models
          • Creating a composite model
        • TensorRT models
          • TensorRT model conversion
      • Deployed model location
    • Inspections
      • Inspection mode
      • Confidence score
      • Custom pipelines
        • Creating a custom pipeline
      • Rules
        • Alert messages
      • Privacy
      • Triggers
        • Input triggers
        • Visual triggers
        • Trigger orchestration
      • Reviewing inspections
    • Optical character recognition
      • Changing the OCR model language
      • Deploying the OCR model
      • Configuring an inspection for OCR
  • Managing
    • Managing your profile
      • Changing your password
      • Updating the API key
    • Managing users
      • Version 8.7 and later
      • Version 8.6
    • Managing data
      • Data retention policy
      • Data transfer policy
        • Updating the data transfer policy
    • Backing up and purging data
    • Edge manager
    • Single sign-on
  • Integrating with Maximo Monitor
    • Assigning a unique prefix to an edge node
    • Connecting the applications
    • Setting up device type metrics and dimensions
    • Creating a device twin
    • Viewing the device configuration
    • Completing the integration
  • Troubleshooting
    • Edge manager diagnostics
    • Container logs
    • Configuring log information
• Notices