---
license: other
license_name: bria-rmbg-1.4
license_link: https://huggingface.co/briaai/RMBG-1.4
library_name: coreml
pipeline_tag: image-segmentation
base_model: briaai/RMBG-1.4
base_model_relation: quantized
tags:
  - coreml
  - core-ml
  - ios
  - macos
  - apple
  - on-device
  - background-removal
  - salient-object-detection
---

# RMBG-1.4 — Core ML

*BRIA AI, 2023*

High-quality background removal. Outputs foreground with alpha mask. 1024×1024 input.

<p><img src="https://huggingface.co/mlboydaisuke/RMBG-1.4-CoreML/resolve/main/media/e731094399.jpg" alt="RMBG-1.4 demo" width="49%"> <img src="https://huggingface.co/mlboydaisuke/RMBG-1.4-CoreML/resolve/main/media/bcd89e2953.jpg" alt="RMBG-1.4 demo" width="49%"></p>

Core ML conversion of [briaai/RMBG-1.4](https://huggingface.co/briaai/RMBG-1.4) for on-device inference on iPhone, iPad and Mac. Converted with `coremltools`; the packages are stateless, so all sequencing and buffering lives in your Swift code.

| | |
|---|---|
| Task | image segmentation |
| Upstream | [briaai/RMBG-1.4](https://huggingface.co/briaai/RMBG-1.4) |
| Packages | 1 |
| Download size | 37 MB |
| Minimum iOS | 17.0 |
| Peak RAM | ~300 MB |

## Files

| File | Size | Compute units | SHA-256 |
|---|---:|---|---|
| `RMBG_1_4.mlpackage.zip` | 37 MB | `cpuOnly` | `a80dbb5f04c922a8…` |
| **Total** | **37 MB** | | |

`compute_units` is not a suggestion -- it is the configuration the conversion was verified against. Moving a package to a different compute unit can silently change the numerics (FP16 attention overflow) or crash on the GPU.

## Download

```bash
hf download mlboydaisuke/coreml-zoo --include "rmbg/*" --local-dir ./rmbg_1_4
unzip './rmbg_1_4/rmbg/*.zip' -d ./rmbg_1_4
```

## Use in Swift

```swift
import CoreML

let config = MLModelConfiguration()
config.computeUnits = .cpuOnly   // as converted — see the table above

// Unzip the .mlpackage, drop it into your Xcode target and Xcode compiles it
// at build time:
let model = try RMBG_1_4(configuration: config)

// ...or compile a downloaded .mlpackage at runtime:
let compiled = try await MLModel.compileModel(at: mlpackageURL)
let model = try MLModel(contentsOf: compiled, configuration: config)
```

## Demo

- **Sample app** — [`sample_apps/RMBGDemo`](https://github.com/john-rocky/CoreML-Models/tree/master/sample_apps/RMBGDemo), a standalone SwiftUI project.
- **Models Zoo** — this model is downloadable and runnable inside the [Models Zoo app](https://apps.apple.com/app/id6762083207) on the App Store, no build required.

## Conversion

- Script: [`convert_rmbg.py`](https://github.com/john-rocky/CoreML-Models/blob/master/conversion_scripts/convert_rmbg.py)
- Pitfalls hit during conversion (FP16 overflow, ANE buffer limits, stride handling): [`docs/coreml_conversion_notes.md`](https://github.com/john-rocky/CoreML-Models/blob/master/docs/coreml_conversion_notes.md)
- Model index: [CoreML-Models](https://github.com/john-rocky/CoreML-Models)

## License

The conversion inherits the upstream license: **Bria RMBG-1.4 License**.
See [https://huggingface.co/briaai/RMBG-1.4](https://huggingface.co/briaai/RMBG-1.4).

> Non-commercial without a separate agreement with Bria.

## Credits

- Upstream authors: [briaai/RMBG-1.4](https://huggingface.co/briaai/RMBG-1.4), 2023
- Core ML conversion: john-rocky (Daisuke Majima)