Universalimagequalityindex

---
name: universalimagequalityindex
description: Compute the UniversalImageQualityIndex metric — provided by torchmetrics. Use when the user has predictions and ground-truth and needs to compute UniversalImageQualityIndex, or asks how to score with UniversalImageQualityIndex.
metadata:
  skill_kind: metric
  source_lib: torchmetrics
  import_path: torchmetrics.UniversalImageQualityIndex
  source: library_introspection
---

# universalimagequalityindex

> Metric `UniversalImageQualityIndex` from `torchmetrics` (torchmetrics.UniversalImageQualityIndex)

## When to invoke this skill

The user has predictions + ground truth and asks to evaluate with UniversalImageQualityIndex, or
mentions `torchmetrics.UniversalImageQualityIndex` directly, or wants the standard torchmetrics implementation.

## Reference signature

```python
from torchmetrics import UniversalImageQualityIndex

# _UniversalImageQualityIndex(kernel_size: collections.abc.Sequence[int] = (11, 11), sigma: collections.abc.Sequence[float] = (1.5, 1.5), reduction: Literal['elementwise_mean', 'sum', 'none', None] = 'elementwise_mean', **kwargs: Any) -> None
```

## Library docstring

```
Wrapper for deprecated import.

>>> import torch
>>> preds = torch.rand([16, 1, 16, 16])
>>> target = preds * 0.75
>>> uqi = _UniversalImageQualityIndex()
>>> uqi(preds, target)
tensor(0.9216)
```

## Quick recipe

```python
import torchmetrics as _m
score = _m.UniversalImageQualityIndex(y_true, y_pred)
```

## Don'ts

- Don't reimplement when the library version handles edge cases (NaN, ties, empty inputs) better than a hand-rolled formula.
- Always check the library version's argument order — sklearn is `(y_true, y_pred)` while torchmetrics is `(preds, target)`.