🏷️ Recognize Anything & Tag2Text

Official PyTorch Implementation of Recognize Anything: A Strong Image Tagging Model and Tag2Text: Guiding Vision-Language Model via Image Tagging.

• Recognize Anything Model(RAM) is an image tagging model, which can recognize any common category with high accuracy.
• Tag2Text is a vision-language model guided by tagging, which can support caption, retrieval and tagging.

Both Tag2Text and RAM exihibit strong recognition ability. We have combined Tag2Text and RAM with localization models (Grounding-DINO and SAM) and developed a strong visual semantic analysis pipeline in the Grounded-SAM project.

💡 Highlight of RAM

RAM is a strong image tagging model, which can recognize any common category with high accuracy.

• Strong and general. RAM exhibits exceptional image tagging capabilities with powerful zero-shot generalization;
  • RAM showcases impressive zero-shot performance, significantly outperforming CLIP and BLIP.
  • RAM even surpasses the fully supervised manners (ML-Decoder).
  • RAM exhibits competitive performance with the Google tagging API.
• Reproducible and affordable. RAM requires Low reproduction cost with open-source and annotation-free dataset;
• Flexible and versatile. RAM offers remarkable flexibility, catering to various application scenarios.

(Green color means fully supervised learning and Blue color means zero-shot performance.)

RAM significantly improves the tagging ability based on the Tag2text framework.

• Accuracy. RAM utilizes a data engine to generate additional annotations and clean incorrect ones, higher accuracy compared to Tag2Text.
• Scope. RAM upgrades the number of fixed tags from 3,400+ to 6,400+ (synonymous reduction to 4,500+ different semantic tags), covering more valuable categories. Moreover, RAM is equipped with open-set capability, feasible to recognize tags not seen during training

🌅 Highlight of Tag2text

Tag2Text is an efficient and controllable vision-language model with tagging guidance.

• Tagging. Tag2Text recognizes 3,400+ commonly human-used categories without manual annotations.
• Captioning. Tag2Text integrates tags information into text generation as the guiding elements, resulting in more controllable and comprehensive descriptions.
• Retrieval. Tag2Text provides tags as additional visible alignment indicators for image-text retrieval.

✍️ TODO

• Release Tag2Text demo.
• Release checkpoints.
• Release inference code.
• Release RAM demo and checkpoints.
• Release training codes.
• Release training datasets.

🧰 Checkpoints

	Name	Backbone	Data	Illustration	Checkpoint
1	RAM-14M	Swin-Large	COCO, VG, SBU, CC-3M, CC-12M	Provide strong image tagging ability.	Download link
2	Tag2Text-14M	Swin-Base	COCO, VG, SBU, CC-3M, CC-12M	Support comprehensive captioning and tagging.	Download link

🏃 Model Inference

Setting Up

1. Install the dependencies::

pip install -r requirements.txt

2. Download RAM pretrained checkpoints.

3. (Optional) To use RAM and Tag2Text in other projects, better to install recognize-anything as a package:

pip install -e .

Then the RAM and Tag2Text model can be imported in other projects:

from ram.models import ram, tag2text

RAM Inference

Get the English and Chinese outputs of the images:

python inference_ram.py  --image images/demo/demo1.jpg 

--pretrained pretrained/ram_swin_large_14m.pth

RAM Inference on Unseen Categories (Open-Set)

Firstly, custom recognition categories in build_openset_label_embedding, then get the tags of the images:

python inference_ram_openset.py  --image images/openset_example.jpg 

--pretrained pretrained/ram_swin_large_14m.pth

Tag2Text Inference

Get the tagging and captioning results:

python inference_tag2text.py  --image images/demo/demo1.jpg 

--pretrained pretrained/tag2text_swin_14m.pth

Or get the tagging and sepcifed captioning results (optional):

python inference_tag2text.py  --image images/demo/demo1.jpg 

--pretrained pretrained/tag2text_swin_14m.pth 

--specified-tags "cloud,sky"

Batch Inference and Evaluation

We release two datasets OpenImages-common (214 seen classes) and OpenImages-rare (200 unseen classes). Copy or sym-link test images of OpenImages v6 to datasets/openimages_common_214/imgs/ and datasets/openimages_rare_200/imgs.

To evaluate RAM on OpenImages-common:

python batch_inference.py \
  --model-type ram \
  --checkpoint pretrained/ram_swin_large_14m.pth \
  --dataset openimages_common_214 \
  --output-dir outputs/ram

To evaluate RAM open-set capability on OpenImages-rare:

python batch_inference.py \
  --model-type ram \
  --checkpoint pretrained/ram_swin_large_14m.pth \
  --open-set \
  --dataset openimages_rare_200 \
  --output-dir outputs/ram_openset

To evaluate Tag2Text on OpenImages-common:

python batch_inference.py \
  --model-type tag2text \
  --checkpoint pretrained/tag2text_swin_14m.pth \
  --dataset openimages_common_214 \
  --output-dir outputs/tag2text

Please refer to batch_inference.py for more options. To get P/R in table 3 of our paper, pass --threshold=0.86 for RAM and --threshold=0.68 for Tag2Text.

To batch inference custom images, you can set up you own datasets following the given two datasets.

🏌️ Model Training/Finetuning

Tag2Text

At present, we can only open source the forward function of Tag2Text as much as possible. To train/finetune Tag2Text on a custom dataset, you can refer to the complete training codebase of BLIP and make the following modifications:

1. Replace the "models/blip.py" file with the current "tag2text.py" model file;
2. Load additional tags based on the original dataloader.

RAM

The training code of RAM cannot be open-sourced temporarily as it is in the company's process.

✒️ Citation

If you find our work to be useful for your research, please consider citing.

@article{zhang2023recognize,
  title={Recognize Anything: A Strong Image Tagging Model},
  author={Zhang, Youcai and Huang, Xinyu and Ma, Jinyu and Li, Zhaoyang and Luo, Zhaochuan and Xie, Yanchun and Qin, Yuzhuo and Luo, Tong and Li, Yaqian and Liu, Shilong and others},
  journal={arXiv preprint arXiv:2306.03514},
  year={2023}
}

@article{huang2023tag2text,
  title={Tag2Text: Guiding Vision-Language Model via Image Tagging},
  author={Huang, Xinyu and Zhang, Youcai and Ma, Jinyu and Tian, Weiwei and Feng, Rui and Zhang, Yuejie and Li, Yaqian and Guo, Yandong and Zhang, Lei},
  journal={arXiv preprint arXiv:2303.05657},
  year={2023}
}

♥️ Acknowledgements

This work is done with the help of the amazing code base of BLIP, thanks very much!

We want to thank @Cheng Rui @Shilong Liu @Ren Tianhe for their help in marrying RAM/Tag2Text with Grounded-SAM.

We also want to thank Ask-Anything, Prompt-can-anything for combining RAM/Tag2Text, which greatly expands the application boundaries of RAM/Tag2Text.