Mac Track — Computer Vision Lab

Local object detection for Apple Silicon (M1/M2/M3) Macs using MediaPipe and YOLOv8 via OpenCV.

Setup

Open a terminal, navigate to your project folder, and run:

python3 -m venv venv
source venv/bin/activate
pip install --upgrade pip
pip install opencv-python mediapipe ultralytics

Then download the MediaPipe EfficientDet model:

curl -L -o efficientdet.tflite "https://storage.googleapis.com/mediapipe-models/object_detector/efficientdet_lite0/int8/1/efficientdet_lite0.tflite"

Scripts

Script	Description
yolo.py	YOLOv8 object detection on your webcam, GPU-accelerated on Apple Silicon. Supports photo and video capture.
yololive.py	Live webcam feed with ghostly/glitchy pose and object detection overlays.
yolopost.py	Batch processes all videos in input/, adds object and pose overlays, and saves results to output/.
pose.py	Real-time human pose (skeleton) tracking via MediaPipe.
poseobj.py	Real-time multi-object tracking via MediaPipe.
emojihead.py	Overlays an emoji on up to four detected faces in your webcam feed.

Examples - Running

MediaPipe — lightweight CPU pipeline using Google's Tasks API:

python poseobj.py

or 

python emojihead.py

YOLOv8 — high-accuracy detection accelerated on Apple Silicon GPU via Metal Performance Shaders:

python yolo.py

or 

python yololive.py

Keyboard Controls

Key	Action
SPACE	Capture a PNG snapshot with tracking boxes
r	Start / stop video recording
q	Quit and close all windows