---
title: Git-RSCLIP
canonical_url: "https://www.modelscope.cn/models/lcybuaa1111/Git-RSCLIP"
md_url: "https://www.modelscope.cn/models/lcybuaa1111/Git-RSCLIP.md"
repository: lcybuaa1111/Git-RSCLIP
last_updated: 2025-03-17
pipeline_tag: image-text-retrieval
tasks:
  - image-text-retrieval
model_type:
  - siglip
architectures:
  - SiglipModel
parameters: 652.2M
tensor_type:
  - F16
library_name:
  - safetensors
  - pytorch
frameworks:
  - Pytorch
downloads: 2770
stars: 1
---

# Git-RSCLIP

> Git-RSCLIP - lcybuaa1111 在 ModelScope 开源的模型。[[Git-RSCLIP]](https://arxiv.org/pdf/2501.00895) is pre-trained on the Git-10M dataset (a global-scale remote sensing image-text pair dataset, consisting of 10 million image-text pairs) at size 256x256, first…

- **Repository**: lcybuaa1111/Git-RSCLIP
- **Tasks**: image-text-retrieval
- **Parameters**: 652.2M
- **Downloads**: 2770
- **Stars**: 1
- **Last updated**: 2025-03-17

Source: https://www.modelscope.cn/models/lcybuaa1111/Git-RSCLIP

---

# Git-RSCLIP

[[Git-RSCLIP]](https://arxiv.org/pdf/2501.00895) is pre-trained on the Git-10M dataset (a global-scale remote sensing image-text pair dataset, consisting of 10 million image-text pairs) at size 256x256, first released in [this repository](https://github.com/chen-yang-liu/Text2Earth). It employs a similar structure to [[google/siglip-large-patch16-256](https://huggingface.co/google/siglip-large-patch16-256)]. 


This is a **large version**, the **base version** is here: [[**Git-RSCLIP-base**](https://huggingface.co/lcybuaa/Git-RSCLIP-base)]

```python
# Download
from modelscope import snapshot_download
model_dir = snapshot_download('lcybuaa1111/Git-RSCLIP')
```

### How to use

#### Use Git-RSCLIP to get image features

```python
from PIL import Image
import requests
from transformers import AutoProcessor, AutoModel
import torch

from modelscope import snapshot_download
model_dir = snapshot_download('lcybuaa1111/Git-RSCLIP')

model = AutoModel.from_pretrained(model_dir)
processor = AutoProcessor.from_pretrained(model_dir)

url = "https://github.com/Chen-Yang-Liu/PromptCC/blob/main/Example/B/train_000051.png?raw=true"
image = Image.open(requests.get(url, stream=True).raw)

inputs = processor(images=image, return_tensors="pt")

with torch.no_grad():
  image_features = model.get_image_features(**inputs)
```


#### zero-shot image classification:

```python
from PIL import Image
import requests
from transformers import AutoProcessor, AutoModel
import torch

from modelscope import snapshot_download
model_dir = snapshot_download('lcybuaa1111/Git-RSCLIP')

model = AutoModel.from_pretrained(model_dir)
processor = AutoProcessor.from_pretrained(model_dir)

url = "https://github.com/Chen-Yang-Liu/PromptCC/blob/main/Example/B/train_000051.png?raw=true"
image = Image.open(requests.get(url, stream=True).raw)

texts = ["a remote sensing image of river", "a remote sensing image of houses and roads"]
inputs = processor(text=texts, images=image, padding="max_length", return_tensors="pt")

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

logits_per_image = outputs.logits_per_image
probs = torch.sigmoid(logits_per_image) # these are the probabilities
top5_indices = torch.argsort(probs, descending=True)[:, :5].cpu().numpy()
top1_indices = top5_indices[:, 0]
print(f"the image 0 is '{top1_indices[0]}'")
```

For more code examples, we refer to the [documentation](https://huggingface.co/transformers/main/model_doc/siglip.html#).


## Training procedure

### Training data

Git-RSCLIP is pre-trained on the Git-10M dataset (a global-scale remote sensing image-text pair dataset, consisting of 10 million image-text pairs) [(Liu et al., 2024)](https://github.com/chen-yang-liu/Text2Earth).

### Preprocessing

Images are resized/rescaled to the same resolution (256x256) and normalized across the RGB channels with mean (0.5, 0.5, 0.5) and standard deviation (0.5, 0.5, 0.5).

Texts are tokenized and padded to the same length (64 tokens).


## Evaluation results

Evaluation of Git-RSCLIP compared to other CLIP is shown below (taken from the paper).

<img src="https://github.com/Chen-Yang-Liu/Text2Earth/blob/main/images/Git-RSCLIP.png?raw=true"
alt="drawing" width="1000"/>

### BibTeX entry and citation info

```bibtex
@misc{liu2025text2earthunlockingtextdrivenremote,
      title={Text2Earth: Unlocking Text-driven Remote Sensing Image Generation with a Global-Scale Dataset and a Foundation Model},
      author={Chenyang Liu and Keyan Chen and Rui Zhao and Zhengxia Zou and Zhenwei Shi},
      year={2025},
      eprint={2501.00895},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2501.00895},
}
```