You can deploy this model and start earning money today!

hugging-mac/
yolov8-seg-coreml

Name: yolov8-seg-coreml
Availability: InStock

instance-segmentation

macos

image-segmentation

agpl-3.0

Views

No views yet

YOLOv8 Instance Segmentation — Core ML

Ready-to-use Core ML conversions of Ultralytics YOLOv8 Seg for instance segmentation on Apple Silicon. The n, s, and m variants are provided as .mlpackage models for use with the CPU, GPU, and Apple Neural Engine.

These models power local computer-vision apps in Hugging Mac, an open-source platform for building AI apps, services, games, plugins, and agents on macOS.

Models

Variant	Package size	Best for
`yolov8n-seg`	7.0 MB	Lowest latency and resource usage
`yolov8s-seg`	23.8 MB	Balanced speed and accuracy
`yolov8m-seg`	54.8 MB	Higher accuracy

Provenance and conversion

Source: Ultralytics assets v8.2.0
Task: COCO 80-class object detection and instance segmentation
Input size: fixed 640 × 640, batch size 1
Precision: FP16 conversion
NMS: not embedded; apply confidence filtering and NMS after inference

Input and output

Name	Type	Shape	Description
`image`	RGB image	`640 × 640`	Letterboxed model input
`predictions`	FP32 multi-array	`1 × 116 × 8400`	Boxes, class scores, and mask coefficients
`prototypes`	FP32 multi-array	`1 × 32 × 160 × 160`	Prototype masks used to reconstruct instance masks

Core ML example

python

from pathlib import Path

import coremltools as ct
from huggingface_hub import snapshot_download
from PIL import Image, ImageOps

root = Path(snapshot_download(
    repo_id="hugging-mac/yolov8-seg-coreml",
    allow_patterns=["yolov8n-seg.mlpackage/**"],
))
model = ct.models.MLModel(
    root / "yolov8n-seg.mlpackage",
    compute_units=ct.ComputeUnit.ALL,
)
image = ImageOps.pad(
    Image.open("image.jpg").convert("RGB"),
    (640, 640),
    color=(114, 114, 114),
)
outputs = model.predict({"image": image})
print(outputs["predictions"].shape)  # (1, 116, 8400)
print(outputs["prototypes"].shape)   # (1, 32, 160, 160)

The outputs are raw. See the Hugging Mac YOLOv8 Seg SDK for complete preprocessing, NMS, mask reconstruction, and coordinate restoration.

Integrity

Package	Directory SHA-256
`yolov8n-seg.mlpackage`	`40e067215baf17be1d4c1a63483538d0fa39c5d0f231cb63248c404e929ce9bc`
`yolov8s-seg.mlpackage`	`ab06bcad128a494510f6451c22cdebb79a58db01f628e9a9a60216b246691399`
`yolov8m-seg.mlpackage`	`685049dbc644c5f9abd088aa398942b052d3256819dfe7bf27456e7136c0e6a5`

License

The converted models retain the upstream Ultralytics licensing terms and are published under AGPL-3.0. Review Ultralytics licensing requirements before commercial or closed-source use. Hugging Mac is not affiliated with or endorsed by Ultralytics.