模型:
nvidia/segformer-b4-finetuned-cityscapes-1024-1024
SegFormer模型在分辨率为1024x1024的CityScapes数据集上进行了微调。该模型在Xie等人的论文( SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers )中提出,并在( this repository )首次发布。
声明:SegFormer发布团队未为该模型撰写模型卡片,因此该模型卡片由Hugging Face团队编写。
SegFormer包括一个分层Transformer编码器和一个轻量级的全MLP解码头部,以实现在ADE20K和Cityscapes等语义分割基准上取得良好的结果。分层Transformer首先在ImageNet-1K上进行预训练,然后添加解码头部并在下游数据集上进行整体微调。
您可以使用原始模型进行语义分割。查看( model hub )以寻找您感兴趣的任务的微调版本。
以下是如何使用该模型将COCO 2017数据集中的图像分类为1000个ImageNet类别之一:
from transformers import SegformerImageProcessor, SegformerForSemanticSegmentation from PIL import Image import requests processor = SegformerImageProcessor.from_pretrained("nvidia/segformer-b4-finetuned-cityscapes-1024-1024") model = SegformerForSemanticSegmentation.from_pretrained("nvidia/segformer-b4-finetuned-cityscapes-1024-1024") url = "http://images.cocodataset.org/val2017/000000039769.jpg" image = Image.open(requests.get(url, stream=True).raw) inputs = processor(images=image, return_tensors="pt") outputs = model(**inputs) logits = outputs.logits # shape (batch_size, num_labels, height/4, width/4)
有关更多代码示例,请参阅( documentation )。
该模型的许可证信息可以在( here )中找到。
@article{DBLP:journals/corr/abs-2105-15203, author = {Enze Xie and Wenhai Wang and Zhiding Yu and Anima Anandkumar and Jose M. Alvarez and Ping Luo}, title = {SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers}, journal = {CoRR}, volume = {abs/2105.15203}, year = {2021}, url = {https://arxiv.org/abs/2105.15203}, eprinttype = {arXiv}, eprint = {2105.15203}, timestamp = {Wed, 02 Jun 2021 11:46:42 +0200}, biburl = {https://dblp.org/rec/journals/corr/abs-2105-15203.bib}, bibsource = {dblp computer science bibliography, https://dblp.org} }