PyKaboo

PyKaboo is a Windows desktop app for synchronized camera acquisition, planner-driven recording, live animal detection, and Arduino TTL control. It supports Basler, FLIR, and USB cameras from one interface and keeps the recording workflow tied to a multi-trial session plan.

Highlights

• Planner-first workflow with trial rows that drive filename, metadata, and recording duration
• Automatic restore of the last planner at startup, including the selected trial
• Auto-advance to the next pending trial when a recording finishes
• Trial row tools for duplicate, copy, paste, move up, and move down
• Live view with optional ROI cropping and frame-drop monitoring
• Live detection panel for RF-DETR Seg and YOLO Seg checkpoints
• Arduino TTL outputs, barcode/sync generation, and live behavior plots
• Recording to MP4 with synchronized metadata exports

A Quick Tour

PyKaboo is built so the operator can stay in one place while moving from setup to acquisition: connect hardware, prepare a trial plan, monitor the live stream, and record with metadata and TTL outputs already aligned.

These views show the app as it is actually used during a session: the trial planner remains central, recording metadata stays visible, and live controls for camera, detection, and Arduino outputs stay close enough to adjust without leaving the acquisition context.

Requirements

• Windows 10 or Windows 11
• Python 3.10 recommended
• ffmpeg available on PATH
• Camera SDKs when using vendor hardware:
• Basler: Pylon SDK plus pypylon
• FLIR machine vision: Spinnaker SDK plus PySpin
• FLIR thermal: flirpy

Install

Conda:

conda env create -f environment.yml
conda activate CamApp

Virtual environment:

python -m venv .venv
.\.venv\Scripts\activate
python -m pip install --upgrade pip
python -m pip install -r requirements.txt

Run

python main.py

Planner Workflow

• Import a CSV plan or build rows directly in the Recording Planner
• The current planner is saved automatically and restored on the next launch
• Select a row to load its metadata into the session form
• Use Ctrl+C and Ctrl+V on planner row selections to copy trial content onto other rows
• Right click planner rows for duplicate, copy, paste, move up, move down, apply, and remove actions
• When a trial finishes recording, PyKaboo marks it as Acquired and selects the next pending row

Outputs

Each recording can produce:

• <name>.mp4
• <name>_metadata.csv
• <name>_metadata.json
• <name>_metadata.txt
• <name>_ttl_states.csv
• <name>_ttl_counts.csv
• <name>_behavior_summary.csv

Arduino Setup

PyKaboo supports:

• StandardFirmata for generic TTL monitoring and output control
• StandardFirmataBarcode for the custom barcode/sync workflow included in StandardFirmataBarcode

Build A Windows EXE

The repo includes:

• camApp-live-detection.spec
• scripts/build_release.ps1

Example:

python -m pip install -r requirements.txt pyinstaller
.\scripts\build_release.ps1 -Version v2026.04.12 -PythonExe python -Clean

Troubleshooting

• ffmpeg not found: add FFmpeg to PATH and restart the shell
• PySpin import errors: use a Spinnaker-compatible wheel and keep numpy<2
• No live inference output: verify the checkpoint path and required ML packages are installed
• Vendor camera connection issues: confirm the camera opens in the vendor SDK viewer first