---
license: other
license_name: cityscapes-non-commercial
license_link: https://www.cityscapes-dataset.com/license/
library_name: libreyolo
pipeline_tag: image-segmentation
tags:
  - image-segmentation
  - semantic-segmentation
  - ppliteseg
  - stdc
  - cityscapes
datasets:
  - cityscapes
---

# LibrePPLiteSegt75-sem

> **NON-COMMERCIAL WEIGHTS.** These weights are trained on Cityscapes. The
> [Cityscapes license](https://www.cityscapes-dataset.com/license/) permits
> distributing abstract derivative models from which the dataset cannot be
> recovered, and restricts the dataset **and its derivatives** to
> non-commercial use. The restriction applies to this checkpoint, not to
> LibreYOLO's MIT code or to the PP-LiteSeg architecture. Weights you train
> from scratch on your own data carry none of it; a fine-tune started from
> this checkpoint inherits it. Read the restriction before you download.

PP-LiteSeg t75 (STDC1 backbone, native 768x1536 canvas), a real-time
semantic segmentation model for Cityscapes' 19 classes, repackaged for
[LibreYOLO](https://github.com/LibreYOLO/libreyolo).

```python
from libreyolo import LibreYOLO

model = LibreYOLO("LibrePPLiteSegt75-sem.pt")
result = model.predict("street.jpg")[0]
mask = result.semantic_mask.data          # (H, W) class IDs on the original canvas
```

This is a genuinely rectangular model: it runs at 768x1536 (height x width),
not a square canvas. Published Cityscapes validation mIoU for the source
checkpoint is 77.56. LibreYOLO has not re-measured that number; it is quoted
from the source release, and exact raw-logit parity does not by itself
reproduce an end-to-end dataset evaluation.

## Source

Derived from [Deci-AI/super-gradients](https://github.com/Deci-AI/super-gradients)
at commit `63de22c404d5740f34f7706c302b37fce3c8fe5d`.
Copyright (c) 2021-2024 Deci AI. Licensed under the Apache License 2.0.

Source artifact: [`pp_lite_t_seg75_cityscapes.pth`](https://d2gjn4b69gu75n.cloudfront.net/models/pp_lite_t_seg75_cityscapes.pth),
SHA-256 `1fdd809572a1b3168727ed0dea32da287c9917ed0ebbfdf8ecab87a3116733f6` (verified before conversion).

The STDC backbone lineage comes from
[MichaelFan01/STDC-Seg](https://github.com/MichaelFan01/STDC-Seg) at commit
`59ff37fbd693b99972c76fcefe97caa14aeb619f`, MIT.
Copyright (c) 2021 Mingyuan Fan. The architecture was cross-checked against
[PaddlePaddle/PaddleSeg](https://github.com/PaddlePaddle/PaddleSeg) at commit
`3c4db66de1d9d59d0628ed87590b6308a2f4aa2a`, Apache-2.0; no PaddleSeg code was
copied.

Paper: [PP-LiteSeg: A Superior Real-Time Semantic Segmentation Model](https://arxiv.org/abs/2204.02681).

## Modifications

State-dict key remapping only: the upstream `net` payload with exactly one
`module.` DDP prefix stripped, wrapped in LibreYOLO v1.0 checkpoint metadata.
Learned parameters are unchanged, and the three training auxiliary heads are
retained so the checkpoint stays trainable. The port reproduces the pinned
upstream exactly (`max_abs_diff == 0.0` on the main logits). See
`weights/convert_ppliteseg_weights.py` and `weights/parity_ppliteseg.py` in the
[LibreYOLO source repository](https://github.com/LibreYOLO/libreyolo).

## License

The **code** is Apache-2.0 (super-gradients) with MIT STDC lineage; both
license texts are in [`LICENSE`](./LICENSE), and attribution is in
[`NOTICE`](./NOTICE).

The **weights in this repository are non-commercial** under the Cityscapes
dataset terms linked above. Downstream users are responsible for complying with
them.