fovea

oaslananka/fovea

★ 0 stars Python 💻 Code/Dev Tools Updated today

Fovea — local-first, edge-AI computer-vision workbench. MCP server (Python) + VS Code extension covering the full YOLO lifecycle: dataset → train → eval → export → benchmark.

View on GitHub →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "fovea": {
      "command": "uvx",
      "args": [
        "fovea"
      ]
    }
  }
}

Or install with pip: pip install fovea

README Excerpt

Fovea is a local-first edge-AI workbench for the full YOLO lifecycle: inspect a dataset, start a run, watch metrics, diagnose failures, export a model, and benchmark the artifact without sending project data to a hosted control plane. This monorepo contains: - `fovea-mcp`: the Python MCP server that exposes dataset, training, evaluation, export, quantization, inference, benchmarking, and management tools.

Tools (20)

annotation_quality_checkbenchmark_latencydataset_convertdataset_find_duplicatesdataset_inspectdataset_splitdataset_validateeval_compareeval_error_analysiseval_per_classeval_runexport_onnxexport_tflitefovea_doctorinfer_batchinfer_imageinfer_rtspmodel_listmodel_profilequantize_int8

Topics

computer-visionedge-aifastmcpmcpml-opsmodel-context-protocolobject-detectiononnxpythontflitetypescriptultralyticsvscode-extensionyolo

fovea

Quick Install

README Excerpt

Tools (20)

Topics

Related Code/Dev Tools Servers