Mask2former Swin Small Cityscapes Semantic

🚀 Mask2Former

The Mask2Former model is trained on Cityscapes semantic segmentation (small - sized version, Swin backbone). It offers a unified approach to handle instance, semantic, and panoptic segmentation tasks.

🚀 Quick Start

The Mask2Former model was introduced in the paper Masked - attention Mask Transformer for Universal Image Segmentation and first released in this repository. Note that the team releasing Mask2Former did not write a model card for this model, and this card is written by the Hugging Face team.

✨ Features

• Unified Segmentation Paradigm: Mask2Former addresses instance, semantic, and panoptic segmentation using the same approach, treating all three tasks as instance segmentation by predicting a set of masks and corresponding labels.
• Performance and Efficiency: It outperforms the previous SOTA, MaskFormer, both in terms of performance and efficiency. It achieves this by replacing the pixel decoder with a multi - scale deformable attention Transformer, adopting a Transformer decoder with masked attention, and improving training efficiency by calculating the loss on subsampled points.

📚 Documentation

Model Description

Mask2Former uses a unified paradigm to handle instance, semantic, and panoptic segmentation. It predicts a set of masks and corresponding labels, treating all three tasks as instance segmentation. It outperforms MaskFormer through several improvements:

• Replacing the pixel decoder with a more advanced multi - scale deformable attention Transformer.
• Adopting a Transformer decoder with masked attention to boost performance without additional computation.
• Improving training efficiency by calculating the loss on subsampled points instead of whole masks.

Intended Uses & Limitations

You can use this particular checkpoint for panoptic segmentation. Check the model hub to find other fine - tuned versions for tasks that interest you.

How to Use

Here is how to use this model:

import requests
import torch
from PIL import Image
from transformers import AutoImageProcessor, Mask2FormerForUniversalSegmentation


# load Mask2Former fine-tuned on Cityscapes semantic segmentation
processor = AutoImageProcessor.from_pretrained("facebook/mask2former-swin-small-cityscapes-semantic")
model = Mask2FormerForUniversalSegmentation.from_pretrained("facebook/mask2former-swin-small-cityscapes-semantic")

url = "http://images.cocodataset.org/val2017/000000039769.jpg"
image = Image.open(requests.get(url, stream=True).raw)
inputs = processor(images=image, return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)

# model predicts class_queries_logits of shape `(batch_size, num_queries)`
# and masks_queries_logits of shape `(batch_size, num_queries, height, width)`
class_queries_logits = outputs.class_queries_logits
masks_queries_logits = outputs.masks_queries_logits

# you can pass them to processor for postprocessing
predicted_semantic_map = processor.post_process_semantic_segmentation(outputs, target_sizes=[image.size[::-1]])[0]
# we refer to the demo notebooks for visualization (see "Resources" section in the Mask2Former docs)

For more code examples, refer to the documentation.

📄 License

The license for this model is other.