Skip to main content

Agent Tools

The following is a list of tools that Clarity AI has access to on our platform.

Workspace & Data Management​

  • Create Workspace - Set up a new workspace with specified name and optional description
  • Add Files to Workspace - Import images, folders, or shapefiles into an existing workspace
  • Get Workspace Information - Retrieve detailed information about workspaces and their contents

Data Analysis (best utilized in the Spectral Explorer)​

  • Get Image Details - Retrieve comprehensive metadata and relationship information for images
  • Process Images - Visualize and apply various operations like cropping, filtering, clustering, etc. Available for most images in the spectral explorer, including:
    • RGB image of the HSI
    • Band Math results
    • Inference results
    • External results
  • Calculate Spectral Indices - Generate and run band math calculations like NDVI or custom spectral indices

Spectral Analysis & Visualization​

  • Plot Spectral Signatures - Generate plots of spectral signatures from labeled image data
  • Plot Individual Signatures - Create plots of single spectral signatures from specific labels
  • Compare Spectral Signatures - Analyze differences between spectral signatures within or across classes
  • Plot Signatures by Class - Visualize spectral signatures organized by ML class categories
  • Plot Reference Spectra - Display reference spectra from scientific spectral libraries
  • Compare with Reference - Match spectral signatures against reference endmember spectra
  • Get Spectral Indices - Access available spectral indices for calculations
  • Search Reference Spectra - Find reference signatures across all available spectral libraries
  • Generate PCA Analysis - Create Principal Component Analysis plots for image layers
  • Analyze Endmembers - Generate endmember analysis visualizations for spectral unmixing

Labeling & Annotation​

  • Auto-Generate Segments - Run Segment Anything to automatically segment objects in images
  • View Segmentation Results - Retrieve AI-generated segmentation results as images
  • Convert Segments to Labels - Transform segmentation results into usable labels with filtering options
  • Search ML Classes - Locate ML class collections and individual classes by name
  • Create ML Classes - Create new ML classes
  • Assign Labels to Classes - Organize labels into ML classes with specified properties
  • Delete Labels - Remove unwanted labels using their identifiers

Dataset Creation & Management​

  • Get Dataset Information - Access metadata for dataset versions and their relationships
  • Create New Dataset - Build new ML datasets with specified processing levels and wavelengths
  • Add Data to Dataset - Include compatible images in dataset versions
  • Find Compatible Images - Identify images with similar characteristics suitable for combined analysis
  • Update Dataset Settings - Modify dataset configuration including sampling methods and preprocessing
  • Clone Dataset Version - Create a new draft by copying an existing dataset version
  • Finalize Dataset - Convert dataset to optimized format ready for machine learning training
  • Analyze Dataset - Get comprehensive statistics and analysis for dataset versions

Model Creation & Management​

  • Get Model Information - Obtain metadata for ML model versions and training details
  • Create New Model - Build new ML models with specified architecture and class configurations
  • Clone Model Version - Create new draft by copying existing model version
  • Update Model Settings - Modify model configuration including hyperparameters and preprocessing
  • Train Model - Start the training process for a model version
  • Analyze Model Performance - Get comprehensive training metrics and performance data
  • Get Model Data Info - Retrieve detailed data information for model versions
  • Get Inference Results Info - Retrieve metadata for model inference outputs
  • Generate Model Reports - Create comprehensive pdf reports for model analysis