birder-project / rope_vit_reg4_b14_capi

Model Card for rope_vit_reg4_b14_capi

A RoPE ViT b14 image encoder pre-trained using CAPI. This model has not been fine-tuned for a specific classification task and is intended to be used as a general-purpose feature extractor or a backbone for downstream tasks like object detection, segmentation, or custom classification.

Model Details

• Model Type: Image classification and detection backbone

• Model Stats:

  • Params (M): 85.7
  • Input image size: 224 x 224

• Dataset: Trained on a diverse dataset of approximately 15M images, including:

  • iNaturalist 2021 (~3.3M)
  • imagenet-w21-webp-wds (~1.6M random subset)
  • WebVision-2.0 (~1.5M random subset)
  • GLDv2 (~820K random subset of 100 chunks)
  • SA-1B (~220K random subset of 20 chunks)
  • COCO (~120K)
  • NABirds (~48K)
  • Birdsnap v1.1 (~44K)
  • CUB-200 2011 (~11K)
  • The Birder dataset (~7M, private dataset)

• Papers:

  • An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale: https://arxiv.org/abs/2010.11929
  • Rotary Position Embedding for Vision Transformer: https://arxiv.org/abs/2403.13298
  • Vision Transformers Need Registers: https://arxiv.org/abs/2309.16588
  • Cluster and Predict Latent Patches for Improved Masked Image Modeling: https://arxiv.org/abs/2502.08769

Model Usage

Image Embeddings

import birder
from birder.inference.classification import infer_image

(net, model_info) = birder.load_pretrained_model("rope_vit_reg4_b14_capi", inference=True)

# Get the image size the model was trained on
size = birder.get_size_from_signature(model_info.signature)

# Create an inference transform
transform = birder.classification_transform(size, model_info.rgb_stats)

image = "path/to/image.jpeg"  # or a PIL image
(out, embedding) = infer_image(net, image, transform, return_embedding=True)
# embedding is a NumPy array with shape of (1, 768)

Detection Feature Map

from PIL import Image
import birder

(net, model_info) = birder.load_pretrained_model("rope_vit_reg4_b14_capi", inference=True)

# Get the image size the model was trained on
size = birder.get_size_from_signature(model_info.signature)

# Create an inference transform
transform = birder.classification_transform(size, model_info.rgb_stats)

image = Image.open("path/to/image.jpeg")
features = net.detection_features(transform(image).unsqueeze(0))
# features is a dict (stage name -> torch.Tensor)
print([(k, v.size()) for k, v in features.items()])
# Output example:
# [('neck', torch.Size([1, 768, 16, 16]))]

Citation

@misc{dosovitskiy2021imageworth16x16words,
      title={An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale}, 
      author={Alexey Dosovitskiy and Lucas Beyer and Alexander Kolesnikov and Dirk Weissenborn and Xiaohua Zhai and Thomas Unterthiner and Mostafa Dehghani and Matthias Minderer and Georg Heigold and Sylvain Gelly and Jakob Uszkoreit and Neil Houlsby},
      year={2021},
      eprint={2010.11929},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2010.11929}, 
}

@misc{heo2024rotarypositionembeddingvision,
      title={Rotary Position Embedding for Vision Transformer},
      author={Byeongho Heo and Song Park and Dongyoon Han and Sangdoo Yun},
      year={2024},
      eprint={2403.13298},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2403.13298},
}

@misc{darcet2024visiontransformersneedregisters,
      title={Vision Transformers Need Registers}, 
      author={Timothée Darcet and Maxime Oquab and Julien Mairal and Piotr Bojanowski},
      year={2024},
      eprint={2309.16588},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2309.16588}, 
}

@misc{darcet2025clusterpredictlatentpatches,
      title={Cluster and Predict Latent Patches for Improved Masked Image Modeling},
      author={Timothée Darcet and Federico Baldassarre and Maxime Oquab and Julien Mairal and Piotr Bojanowski},
      year={2025},
      eprint={2502.08769},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2502.08769},
}