Skip to content

wildetect-cli.md

Latest commit

 

History

History
343 lines (227 loc) · 9.23 KB

wildetect-cli.md

File metadata and controls

343 lines (227 loc) · 9.23 KB

WildDetect CLI Reference

Complete command-line interface reference for the wildetect CLI.

WildDetect uses Typer with nested subcommand groups. All commands are config-driven — you provide a YAML configuration file rather than many individual flags.

wildetect [COMMAND_GROUP] [COMMAND] [OPTIONS]

Global Options

Option Description
--version Show the package version and exit
--help Show help message and exit

detection — Detection and Analysis Commands

Core commands for running wildlife detection, census campaigns, and analysis.

detection detect

Run wildlife detection on images using AI models.

wildetect detection detect [OPTIONS]
Option Type Default Description
-c, --config PATH None Path to YAML configuration file

The detection command is config-driven. All parameters (model, image paths, processing options, etc.) are specified in the YAML config file.

Example:

wildetect detection detect -c config/detection.yaml

See Detection Config Reference for all configuration fields.

detection census

Run a wildlife census campaign with detection, statistics, geographic analysis, and reporting.

wildetect detection census [OPTIONS]
Option Type Default Description
-c, --config PATH None Path to YAML census configuration file

Example:

wildetect detection census -c config/census.yaml

The census command runs a complete campaign pipeline:

  1. Load images (from directory, paths list, or Label Studio)
  2. Run detection pipeline
  3. Compute species counts and statistics
  4. Export to FiftyOne / Label Studio
  5. Generate geographic visualizations

See Census Config Reference for all configuration fields.

detection analyze

Analyze detection results with geographic and statistical analysis.

wildetect detection analyze RESULTS_PATH [OPTIONS]
Argument / Option Type Default Description
RESULTS_PATH PATH (required) Path to detection results JSON file
-o, --output PATH analysis Output directory for analysis results
-v, --verbose bool false Enable verbose logging

Example:

wildetect detection analyze results/results.json -o analysis_output/

visualization — Visualization Commands

Commands for geographic visualization and GPS coordinate extraction.

visualization visualize

Create geographic visualizations from detection results.

wildetect visualization visualize [OPTIONS]
Option Type Default Description
-c, --config PATH None Path to YAML visualization configuration file
-v, --verbose bool false Enable verbose logging

!!! warning This command is currently not fully implemented (raises NotImplementedError).

See Visualization Config Reference for configuration details.

visualization extract-gps-coordinates

Extract GPS coordinates from images and export detection/annotation reports to CSV.

wildetect visualization extract-gps-coordinates [OPTIONS]
Option Type Default Description
-c, --config PATH None Path to YAML configuration file
-v, --verbose bool false Enable verbose logging

Uses the visualize config type. The detection type can be set to annotations or predictions to control what is exported.

See GPS Extraction Config Reference.

services — Service Management Commands

Commands for launching web interfaces, APIs, and external tools.

services ui

Launch the WildDetect Streamlit web interface.

wildetect services ui [OPTIONS]
Option Type Default Description
-p, --port int 8501 Port to run the UI on
-h, --host str localhost Host to bind to
--no-browser bool true Don't open browser automatically

Example:

wildetect services ui --port 8502

services api

Launch the WildDetect FastAPI REST server.

wildetect services api [OPTIONS]
Option Type Default Description
-h, --host str 0.0.0.0 Host to bind to
-p, --port int 8000 Port to run the API on
--reload bool false Enable auto-reload for development

Example:

wildetect services api --port 8080 --reload

API documentation is available at http://<host>:<port>/docs (Swagger) and http://<host>:<port>/redoc.

services fiftyone

Manage FiftyOne datasets for wildlife detection — launch the app, get dataset info, or export data.

wildetect services fiftyone [OPTIONS]
Option Type Default Description
-d, --dataset str wildlife_detection Dataset name
-a, --action str launch Action: launch, info, or export
-f, --format str coco Export format (coco, yolo, pascal)
-o, --output PATH None Export output path

Examples:

# Launch FiftyOne app
wildetect services fiftyone -a launch

# Get dataset info
wildetect services fiftyone -a info -d my_dataset

# Export dataset
wildetect services fiftyone -a export -d my_dataset -f coco -o exports/

utils — Utility Commands

System utilities, configuration helpers, and maintenance.

utils info

Display system information, dependency status, and environment variables.

wildetect utils info

Shows: Python version, PyTorch version, CUDA availability, GPU info, installed dependencies, project paths, and key environment variables.

utils clear-results

Delete all detection results in a specified directory (with confirmation prompt).

wildetect utils clear-results [OPTIONS]
Option Type Default Description
--results-dir str results Directory containing results to clear

utils install-cuda

Install PyTorch with CUDA support for optimal performance.

wildetect utils install-cuda [OPTIONS]
Option Type Default Description
-c, --cuda-version str None Specific CUDA version (118, 121)
--cpu-only bool false Force CPU-only installation
-v, --verbose bool false Verbose logging

utils create-config

Generate a default YAML configuration file for a command type.

wildetect utils create-config COMMAND_TYPE [OPTIONS]
Argument / Option Type Default Description
COMMAND_TYPE str (required) Command type: detect, census, or visualize
-o, --output PATH None Output path for the config file
--pydantic bool true Use Pydantic models for validation

Example:

# Generate a default detection config
wildetect utils create-config detect -o my_detection_config.yaml

benchmarking — Benchmarking Commands

Performance testing and hyperparameter optimization for detection pipelines.

benchmarking detection

Run detection pipeline benchmarking with Optuna hyperparameter optimization.

wildetect benchmarking detection [OPTIONS]
Option Type Default Description
-c, --config PATH (required) Path to benchmark configuration file

Example:

wildetect benchmarking detection -c config/benchmark.yaml

The benchmark command will:

  1. Load the benchmark configuration from the YAML file
  2. Find test images in the configured directory
  3. Run Optuna optimization to find the best hyperparameters
  4. Save results and optionally generate performance plots

See Benchmark Config Reference for all configuration fields.

Quick Reference

Command Description
wildetect detection detect -c CONFIG Run wildlife detection
wildetect detection census -c CONFIG Run census campaign
wildetect detection analyze RESULTS Analyze detection results
wildetect visualization extract-gps-coordinates -c CONFIG Extract GPS coordinates
wildetect services ui Launch web UI
wildetect services api Launch REST API
wildetect services fiftyone -a launch Launch FiftyOne viewer
wildetect utils info Show system info
wildetect utils create-config TYPE Generate default config
wildetect benchmarking detection -c CONFIG Run benchmarks

For configuration file details, see Configuration Reference.
For shell scripts, see WildDetect Scripts.