Feat/embedding extraction

.gitignore

@@ -2,6 +2,8 @@
 __pycache__/
 *.py[cod]
 *$py.class
+visualizer
+.vscode
 
 # C extensions
 *.so
@@ -227,5 +229,10 @@ docs/cookbooks/*.py
 docs/cookbooks/export_executorch/
 docs/cookbooks/export_tensorrt/
 
+# deprecated code
+visualizer
+
+# VSCode metadata directory
+.vscode
 # MacOS files
 .DS_Store

CHANGELOG.md

@@ -8,6 +8,10 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 
 ### Added
 
+- `RFDETR.predict(..., return_embeddings=True)` now attaches a per-detection embedding vector as `detections.data["embeddings"]` (or `key_points.data["embeddings"]` for keypoint outputs), shape `(K, H)`, gathered with the same indices used for boxes/masks/keypoints — useful for downstream similarity search, clustering, or re-identification. On the eager (unoptimized) model this can be toggled per `predict()` call. On a model optimized via `model.inference(...)`, the exported/traced forward pass has fixed control flow, so `inference()` now also accepts `return_embeddings` and the value must be decided at optimization time and match the `return_embeddings` passed to `predict()`; a mismatch raises `RuntimeError`.
+
+- Added the `embedding-extraction` cookbook, a Colab-runnable notebook that extracts per-detection query embeddings on a public COCO 2017 validation subset (streamed from a Hugging Face parquet export, no API key) and applies them to dataset quality auditing: ranking mislabelled annotations against injected label noise (reported as ROC-AUC and precision@k versus random review), surfacing confident detections with no matching annotation, and nearest-neighbour object retrieval.
+
 - `deploy_to_roboflow()`'s `version` argument is now optional: when omitted, the highest existing dataset version of the target project is resolved automatically via the Roboflow API (falling back to version `1` for a project with no generated versions, where the Roboflow SDK then raises its usual "Version number 1 is not found."). Passing an explicit `version` behaves exactly as before, with no extra API call. ([#1116](https://github.com/roboflow/rf-detr/issues/1116))
 
 - Added a live opt-in end-to-end CI job (`roboflow-deploy-e2e`, `-m e2e_roboflow`) that generates a fresh dataset version in a dedicated Roboflow test project, deploys a real model with `version` omitted, and independently polls the server-side trained-model status — catching silent server-side upload failures that `deploy_to_roboflow()`'s return value cannot surface. ([#1116](https://github.com/roboflow/rf-detr/issues/1116))

docs/cookbooks/NOTES.md

@@ -36,7 +36,7 @@ If jupytext is not installed: `pip install jupytext` (or `uv add jupytext --dev`
     description: One sentence describing what the notebook demonstrates.
 ```
 
-Available labels (reuse these to keep tags standardised): `TRAINING`, `AUGMENTATION`, `EXPORT`, `TFLITE`, `PYTORCH LIGHTNING`, `INFERENCE`, `SEGMENTATION`, `DEPLOY`. Current tag colours are assigned dynamically by the docs UI, so they may change if cards or labels are added or reordered.
+Available labels (reuse these to keep tags standardised): `TRAINING`, `AUGMENTATION`, `EXPORT`, `TFLITE`, `PYTORCH LIGHTNING`, `INFERENCE`, `SEGMENTATION`, `DEPLOY`, `EMBEDDINGS`. Current tag colours are assigned dynamically by the docs UI, so they may change if cards or labels are added or reordered.
 
 For newly added or updated notebooks, write markdown cells in plain, notebook-portable Markdown only — no MkDocs Material syntax (`!!! note`, `=== "Tab"`, admonition blocks). `mkdocs-jupyter` renders the site copy through the same plain renderer as a raw `.ipynb`, so MkDocs-only syntax shows up as literal text (e.g. `!!! warning "..."`) instead of a styled callout. Use a blockquote (`> **Note:** ...`) for callouts instead.
 
@@ -51,6 +51,7 @@ For newly added or updated notebooks, write markdown cells in plain, notebook-po
 | ----------------------------------- | ----------------------------------------------- | ------- |
 | `custom-augmentations.ipynb`        | Custom Augmentations and Live Training Progress | v1.5.0  |
 | `custom-optimizer-scheduler.ipynb`  | Custom Optimizer and LR Scheduler               | v1.9.0  |
+| `embedding-extraction.ipynb`        | Query Embeddings for Dataset Quality Auditing   | v1.10.0 |
 | `export-coreml.ipynb`               | Export to Native CoreML & Run Inference         | v1.9.0  |
 | `export-tensorrt.ipynb`             | Export to TensorRT & Run Inference              | v1.9.0  |
 | `export-executorch.ipynb`           | Export to ExecuTorch & Run Inference            | v1.9.0  |

docs/cookbooks/cards.yaml

@@ -1,4 +1,10 @@
 cards:
+  - href: embedding-extraction/
+    name: "Query Embeddings for Dataset Quality Auditing"
+    labels: [INFERENCE, EMBEDDINGS]
+    version: v1.10.0
+    author: unaxEtxeberriaBieleDigital
+    description: "Extract per-detection query embeddings with predict(return_embeddings=True) and use them to rank mislabelled annotations, surface unannotated objects, and retrieve similar instances — measured on a public COCO subset."
   - href: export-coreml/
     name: "Export to Native CoreML & Run Inference"
     labels: [EXPORT, INFERENCE, DEPLOY]