NearID produces identity-aware image embeddings that remain stable across background and context changes while correctly rejecting near-identity distractors — visually similar but different instances placed in the same context. It is designed for evaluating identity preservation in personalized image generation.

pip install nearid
# or install from source:
# pip install -e .

from transformers import AutoModel, AutoImageProcessor
from PIL import Image

model = AutoModel.from_pretrained("Aleksandar/nearid-siglip2", trust_remote_code=True)
processor = AutoImageProcessor.from_pretrained("Aleksandar/nearid-siglip2")

inputs = processor(images=Image.open("photo.jpg"), return_tensors="pt")
embedding = model.get_image_features(**inputs)  # [1, 1152], L2-normalised

SSR and PA are averaged across seven inpainting settings (three excluded from training). MO/MOpair = metric-to-oracle correlation; MH = metric-to-human correlation (Fisher-z averaged).

Inference only:

pip install -e .

Training & evaluation:

conda env create -f environment.yaml
conda activate nearid
pip install -e ".[all]"

import torch

emb_a = model.get_image_features(**processor(images=img_a, return_tensors="pt"))
emb_b = model.get_image_features(**processor(images=img_b, return_tensors="pt"))

similarity = (emb_a @ emb_b.T).item()  # cosine similarity

images = [Image.open(p) for p in image_paths]
inputs = processor(images=images, return_tensors="pt", padding=True)
embeddings = model.get_image_features(**inputs)  # [B, 1152]

sim_matrix = embeddings @ embeddings.T

Train with the NearID loss (extended InfoNCE with near-identity distractor ranking):

accelerate launch -m training.train \
    --loss_config "infonce_ext:1.0" \
    --head_type map --head_out_dim 1152 \
    --lr 1e-4 --epochs 11 --data.batch_size 128

See docs/TRAINING.md for the full guide.

# Step 1: compute similarities
python -m evaluation.sim_test \
    --mode fullneg --model "Aleksandar/nearid-siglip2" \
    --ds "Aleksandar/NearID" --ds_neg "path/to/negatives" \
    --output_folder runs/evals/ --batch_size 64

# Step 2: aggregate tables
python -m evaluation.gen_tables --root runs/evals/ --overlap primary

See docs/EVALUATION.md for the full guide.

The NearID benchmark consists of multi-view positives and near-identity distractors generated by an ensemble of inpainting pipelines. All datasets are released under CC-BY-4.0.

from datasets import load_dataset

positives = load_dataset("Aleksandar/NearID")
negatives = load_dataset("Aleksandar/NearID-Flux")

@article{cvejic2026nearid,
  title={NearID: Identity Representation Learning via Near-identity Distractors},
  author={Cvejic, Aleksandar and Abdal, Rameen and Eldesokey, Abdelrahman and Ghanem, Bernard and Wonka, Peter},
  journal={arXiv preprint arXiv:2604.01973},
  year={2026}
}

This work was supported by King Abdullah University of Science and Technology (KAUST) and Snap Inc.

• Code & model weights: Apache License 2.0. See LICENSE.
• Datasets: CC-BY-4.0. Derived from SynCD (MIT License).