---
title: HVI-CIDNet-LOLv1-fp32
canonical_url: "https://www.modelscope.cn/models/mlx-community/HVI-CIDNet-LOLv1-fp32"
md_url: "https://www.modelscope.cn/models/mlx-community/HVI-CIDNet-LOLv1-fp32.md"
repository: mlx-community/HVI-CIDNet-LOLv1-fp32
last_updated: 2026-07-28
license: mit
pipeline_tag: image-to-image
tasks:
  - image-to-image
base_model:
  - Fediory/HVI-CIDNet
base_model_relation: finetune
parameters: 2.0M
tensor_type:
  - F32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 11
stars: 0
tags:
  - mlx
  - low-light-enhancement
  - exposure-correction
  - image-enhancement
  - hvi-cidnet
---

# HVI-CIDNet-LOLv1-fp32

> HVI-CIDNet-LOLv1-fp32 - mlx-community 在 ModelScope 开源的模型。mlx-community/HVI-CIDNet-LOLv1-fp32

mlx-community/HVI-CIDNet-LOLv1-fp32 是 ModelScope 魔搭社区上的 2.0M 参数image-to-image模型，采用 mit 许可，基于 Fediory/HVI-CIDNet 构建。

- **Repository**: mlx-community/HVI-CIDNet-LOLv1-fp32
- **License**: mit
- **Tasks**: image-to-image
- **Parameters**: 2.0M
- **Base model**: Fediory/HVI-CIDNet
- **Tags**: mlx, low-light-enhancement, exposure-correction, image-enhancement, hvi-cidnet
- **Downloads**: 11
- **Stars**: 0
- **Last updated**: 2026-07-28

Source: https://www.modelscope.cn/models/mlx-community/HVI-CIDNet-LOLv1-fp32

---

# mlx-community/HVI-CIDNet-LOLv1-fp32

[HVI-CIDNet](https://github.com/Fediory/HVI-CIDNet) low-light / exposure correction, converted to
**Apple MLX** for Apple-Silicon inference via the
[`mlx-cidnet-swift`](https://github.com/xocialize/mlx-cidnet-swift) Swift package.

Yan et al., *HVI: A New Color Space for Low-light Image Enhancement*, **CVPR 2025** — 1st place,
NTIRE 2025 Low-Light Enhancement Challenge. 1,975,569 parameters (7.9 MB). Underexposed image in →
re-exposed image at the same resolution out.

The headline LOL-v1 model (perceptual loss). Strongest lift on *severely* underexposed input (+14.50 dB at 0.15x exposure) and the best published LOL numbers.

## Use with mlx-cidnet-swift

```swift
import CIDNetMLXCore

let model = CIDNet()
try model.loadWeights(from: weightsURL)   // model.safetensors from this repo
let brightened = model(imageNHWC)          // NHWC RGB in [0,1]
```

Or as an MLXEngine `imageRelight` ModelPackage (`MLXCIDNet.CIDNetRelightPackage`), which resolves
this repo via the Hub and applies the exposure gate below automatically.

## ⚠️ Gate this model on input exposure — it is not safe to apply unconditionally

Every published checkpoint drives its output toward a **target mean luma regardless of input**. On
an image that is already correctly exposed the model therefore *degrades* it. Measured against a
correctly-exposed reference:

| checkpoint | applied to an already-correct exposure |
|---|---|
| Generalization | 23.37 dB |
| LOLv1-wperc | 20.99 dB |
| SICE | **16.10 dB** |

SICE's output mean stays within 0.35–0.40 across a **12× input range** — it is an auto-exposure
normalizer, not a shadow lift.

Estimate input luma and **bypass above a threshold**; expose a strength blend for partial
application. `mlx-cidnet-swift` does both. Note also that Generalization scores **−0.52 dB — worse
than doing nothing** — at 0.70× exposure, which is the *moderate* underexposure regime that
LOL-trained models are documented to handle badly.

## Conversion

MLX **NHWC** layout: 191 tensors → 81 conv + 61 depthwise transposed `(O,I,kH,kW) → (O,kH,kW,I)`,
49 passthrough. Upstream already publishes safetensors, so this is a relayout, not a format change.

`trans.density_k` is a **learned** parameter and differs per checkpoint (this one: **1.1255**; init is
0.2). It rides in the weights — do not hardcode it.

## Parity

Gated against the PyTorch oracle on the CPU stream, fp32, judged on relative error:

- **key contract** — 191 tensors / 1,975,569 params / 0 missing / 0 unused, strict load
- **HVI colour transform** — **bit-exact** (worst 6.6e-07), including the degenerate colours and
  the yellow/cyan channel *ties* that decide the masking priority
- **primitives** — PReLU exactly 0.00e+00; the align-corners resamplers ~5e-07
- **full model** — cosine **1.00000000** at 64² / 128² / 256², and on a severely underexposed tile

Weights: MIT (Fediory/HVI-CIDNet, published first-party by the authors). Port code: MIT.