license: apache-2.0
pipeline_tag: feature-extraction
tags:
- image-text-retrieval
- jepa
- contrastive-learning
- video-encoder
- pytorch
scout-eccv-hard-neg
Part of SCOUT (Sim-to-real Contrastive Outlier Understanding via Text), an ECCV 2026 workshop paper on JEPA-based text-based person re-identification for the AI City Challenge Track 4 (Sim2Real). Paper: SCOUT: Sim-to-Real Text-Based Person Retrieval by Embedding-Space Prediction over Frozen Video Features. Code: https://github.com/abtraore/SCOUT-ECCV
v17 recipe + pre-mined hard-negative pairs (CMP identity hard-negs) guaranteed in every batch.
- Val mAP@10: 0.879
- Val Hit@10: n/a
- Trainable / total parameters: 549.6M / 1161.8M
A fusion member of the paper's headline board system.
What's in this repo
Only the parameters SCOUT actually trained: a model.safetensors (trainable predictor weights, and
for ExPLoRA/LoRA variants, the adapted encoder deltas) plus a config.json recording the exact recipe.
Frozen backbone weights (V-JEPA2 from Meta, the text encoder from its own source) are not included
here; they're downloaded from their own public sources at load time.
Usage
from scripts.load_from_hf import load_from_hf
model = load_from_hf("Abdrah/scout-eccv-hard-neg").eval()
# model.encode_image(pixel_values) # [B, 3, 384, 384] in [0, 1] -> [B, out_dim] L2-normalized
# model.encode_text(["a caption"]) # -> [B, out_dim] L2-normalized
Or drop it straight into the repo's own evaluation script:
uv run python scripts/local_eval.py --hf-repo Abdrah/scout-eccv-hard-neg --pab-root /path/to/PAB --query-source val
License
Apache 2.0 (see the source repo's LICENSE/NOTICE). This checkpoint does not redistribute any frozen backbone weights, which carry their own upstream licenses.