README.md

---
pipeline_tag: image-segmentation
license: mit
base_model: tue-mps/coco_panoptic_eomt_large_640
library_name: kerasformers
tags:
- keras
- kerasformers
- eomt
- panoptic-segmentation
- image-segmentation
- arxiv:2503.19108
- pytorch
- jax
- tf
---

## ***See [our collection](https://huggingface.co/collections/kerasformers/eomt-6a6a8e0309f2602fa0b94bb8) for all versions of EoMT.***

# Run EoMT with Keras 3: JAX, PyTorch, or TensorFlow

[![GitHub](https://img.shields.io/badge/GitHub-KerasFormers-black?logo=github)](https://github.com/IMvision12/KerasFormers) [![Docs](https://img.shields.io/badge/Docs-EoMT-blue)](https://imvision12.github.io/KerasFormers/eomt/) [![Collection](https://img.shields.io/badge/HF-EoMT%20collection-yellow)](https://huggingface.co/collections/kerasformers/eomt-6a6a8e0309f2602fa0b94bb8)

# kerasformers/eomt_large_coco_panoptic_640

Paper: [Your ViT is Secretly an Image Segmentation Model (arXiv:2503.19108)](https://arxiv.org/abs/2503.19108) · [HF Papers](https://huggingface.co/papers/2503.19108)

EoMT (Encoder-only Mask Transformer) keeps segmentation inside a plain ViT: learned query tokens are concatenated with patch tokens and run through the same ViT blocks. No pixel decoder, no deformable attention decoder.

For more details on the model, please go to the upstream [model card](https://huggingface.co/tue-mps/coco_panoptic_eomt_large_640).

Pure-**Keras 3** conversion of [`tue-mps/coco_panoptic_eomt_large_640`](https://huggingface.co/tue-mps/coco_panoptic_eomt_large_640) for [kerasformers](https://github.com/IMvision12/KerasFormers). One implementation runs unmodified on **TensorFlow / Torch / JAX**.

This is a **panoptic** checkpoint (`EoMTUniversalSegment`).

## ✨ Quick start

```python
import os
os.environ["KERAS_BACKEND"] = "torch"  # or "jax" / "tensorflow"

from PIL import Image
from kerasformers.models.eomt import EoMTUniversalSegment, EoMTImageProcessor

model = EoMTUniversalSegment.from_weights("kerasformers/eomt_large_coco_panoptic_640")
processor = EoMTImageProcessor.from_weights("kerasformers/eomt_large_coco_panoptic_640")

image = Image.open("your_image.jpg").convert("RGB")
output = model(processor(image)["pixel_values"], training=False)
result = processor.post_process_panoptic_segmentation(
    output, target_size=(image.height, image.width)
)
print(result["segmentation"].shape)
```

Load any EoMT variant the same way with `from_weights("kerasformers/<variant>")`:

| Variant | Hub | Task |
|---|---|---|
| `eomt_small_coco_panoptic_640` | [`kerasformers/eomt_small_coco_panoptic_640`](https://huggingface.co/kerasformers/eomt_small_coco_panoptic_640) | panoptic |
| `eomt_base_coco_panoptic_640` | [`kerasformers/eomt_base_coco_panoptic_640`](https://huggingface.co/kerasformers/eomt_base_coco_panoptic_640) | panoptic |
| `eomt_large_coco_panoptic_640` | [`kerasformers/eomt_large_coco_panoptic_640`](https://huggingface.co/kerasformers/eomt_large_coco_panoptic_640) | panoptic |
| `eomt_large_coco_instance_640` | [`kerasformers/eomt_large_coco_instance_640`](https://huggingface.co/kerasformers/eomt_large_coco_instance_640) | instance |
| `eomt_large_ade20k_semantic_512` | [`kerasformers/eomt_large_ade20k_semantic_512`](https://huggingface.co/kerasformers/eomt_large_ade20k_semantic_512) | semantic |

## Tips

- Set `KERAS_BACKEND` **before** importing Keras / kerasformers.
- Use the post-processor that matches the checkpoint task (panoptic / instance / semantic).
- See [EoMT docs]({DOCS_URL}) and [Loading Weights](https://imvision12.github.io/KerasFormers/loading_weights/).
- Community / upstream weights: `EoMTUniversalSegment.from_weights("hf:tue-mps/coco_panoptic_eomt_large_640")`.

## Special Thanks

A huge thank you to the TU/e MPS EoMT authors for creating and releasing these models.

License: MIT.