feat(anomalydino): Introduces anomalydino with semgrep "fixes" #3193

rajeshgangireddy · 2025-12-04T16:44:49Z

📝 Description

AnomalyDino implemented by @alfieroddanintel at : #3105
However, semgrep was failing even after following the safe way of loading models.
Therefore, we apply semgrep exception on top.

✨ Changes

Select what type of change your PR is:

✅ Checklist

Before you submit your pull request, please make sure you have completed the following steps:

📚 I have made the necessary updates to the documentation (if applicable).
🧪 I have written tests that support my changes and prove that my fix is effective or my feature works (if applicable).
🏷️ My PR title follows conventional commit format.

For more information about code review checklists, see the Code Review Checklist.

* init commit Signed-off-by: Alfie Roddan <[email protected]> * add cdist instead of euclidean distance, divide by 2 for cosine and improve comments Signed-off-by: Alfie Roddan <[email protected]> * remove redundant euclidean distance Signed-off-by: Alfie Roddan <[email protected]> * Add AnomalyDINO to model list. Also alphabetically re-order some models Signed-off-by: Alfie Roddan <[email protected]> * add precision modifier Signed-off-by: Alfie Roddan <[email protected]> * remove fit comments. small typo of shape dimensions Signed-off-by: Alfie Roddan <[email protected]> * update docs for anomaly dino Signed-off-by: Alfie Roddan <[email protected]> * cleanup docstrings Signed-off-by: Alfie Roddan <[email protected]> * add unit tests for anomalydino. change distance computation from cdist to matmul, work with half tensors Signed-off-by: Alfie Roddan <[email protected]> * add vit/dino implementation (no xformers). implement factory class for generating dinov2. update anomaly_dino to use factory method Signed-off-by: Alfie Roddan <[email protected]> * change dinov2loader to factory method, remove duplicated components from dinomaly Signed-off-by: Alfie Roddan <[email protected]> * add from_name back Signed-off-by: Alfie Roddan <[email protected]> * add tests for vit and dinov2loader Signed-off-by: Alfie Roddan <[email protected]> * update docstrings Signed-off-by: Alfie Roddan <[email protected]> * fix(accelerator): Adding name method in XPUAccelerator (#3108) * Update xpu.py regarind PR #3092 Added the name method to fix an issue related to a newly added feature in lightning 2.5.6 Signed-off-by: Niclas <[email protected]> * Update xpu.py Signed-off-by: Niclas <[email protected]> * Update xpu.py Signed-off-by: Niclas <[email protected]> * Update xpu.py with docstring Signed-off-by: Niclas <[email protected]> * Update xpu.py with correct docstring Signed-off-by: Niclas <[email protected]> * added name method for XPUAccelerator Signed-off-by: waschsalz <[email protected]> --------- Signed-off-by: Niclas <[email protected]> Signed-off-by: waschsalz <[email protected]> Signed-off-by: Alfie Roddan <[email protected]> * change licesning with meta. Tensor is torch.Tensor. remove __future__. Signed-off-by: Alfie Roddan <[email protected]> * Update src/anomalib/models/components/dinov2/layers/block.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/attention.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/dino_head.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/drop_path.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/layer_scale.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/mlp.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/patch_embed.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/layers/swiglu_ffn.py Signed-off-by: Samet Akcay <[email protected]> * Update src/anomalib/models/components/dinov2/vision_transformer.py Signed-off-by: Samet Akcay <[email protected]> --------- Signed-off-by: Alfie Roddan <[email protected]> Signed-off-by: Niclas <[email protected]> Signed-off-by: waschsalz <[email protected]> Signed-off-by: Samet Akcay <[email protected]> Co-authored-by: Niclas <[email protected]> Co-authored-by: Samet Akcay <[email protected]> Co-authored-by: Samet Akcay <[email protected]> Co-authored-by: Rajesh Gangireddy <[email protected]>

Copilot

Pull request overview

This PR introduces the AnomalyDINO model for anomaly detection using DINOv2 vision transformers. The implementation refactors shared DINOv2 components into a common anomalib.models.components.dinov2 module that can be used by both AnomalyDINO and Dinomaly models. Semgrep exceptions are applied to allow safe model weight loading with weights_only=True.

Key Changes:

Introduces AnomalyDINO model with memory-bank-based anomaly detection
Refactors DINOv2 vision transformer and loader components into shared components module
Migrates Dinomaly to use the new shared DINOv2 components

Reviewed changes

Copilot reviewed 31 out of 31 changed files in this pull request and generated 4 comments.

Show a summary per file

File	Description
`src/anomalib/models/image/anomaly_dino/torch_model.py`	Core PyTorch implementation of AnomalyDINO with feature extraction and kNN-based scoring
`src/anomalib/models/image/anomaly_dino/lightning_model.py`	Lightning wrapper for AnomalyDINO training and inference
`src/anomalib/models/components/dinov2/vision_transformer.py`	Shared DINOv2 vision transformer implementation
`src/anomalib/models/components/dinov2/dinov2_loader.py`	Shared loader for DINOv2 pretrained weights
`src/anomalib/models/components/dinov2/layers/`	Shared DINOv2 layer components (attention, blocks, MLP, etc.)
`src/anomalib/models/image/dinomaly/torch_model.py`	Updated to use shared DINOv2 loader
`src/anomalib/models/image/dinomaly/components/`	Removed duplicated DINOv2 loader and layers, now imports from shared components
`tests/unit/models/image/anomaly_dino/test_torch_model.py`	Unit tests for AnomalyDINO torch model
`tests/unit/models/components/dinov2/`	Unit tests for shared DINOv2 components
`examples/configs/model/anomaly_dino.yaml`	Configuration example for AnomalyDINO
`docs/source/markdown/guides/reference/models/image/anomaly_dino.md`	Documentation for AnomalyDINO model

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

tests/unit/models/image/anomaly_dino/test_torch_model.py

src/anomalib/models/image/anomaly_dino/torch_model.py

src/anomalib/models/components/dinov2/layers/drop_path.py

src/anomalib/models/image/anomaly_dino/lightning_model.py

src/anomalib/models/components/dinov2/dinov2_loader.py

src/anomalib/models/components/dinov2/vision_transformer.py

src/anomalib/models/components/dinov2/dinov2_loader.py

…ading

Copilot

Pull request overview

Copilot reviewed 31 out of 31 changed files in this pull request and generated 5 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2025-12-04T17:00:50Z

src/anomalib/models/image/__init__.py

+    "Dinomaly",
    "Dfm",


The import order in __all__ should maintain alphabetical sorting. Dinomaly appears after Dfm but should come after Dfkde alphabetically. Move Dinomaly to line 79 (between Dfkde and Dfm) to maintain consistency.

Suggested change

"Dinomaly",

"Dfm",

"Dfm",

"Dinomaly",

Copilot · 2025-12-04T17:00:50Z

src/anomalib/models/__init__.py

    "EfficientAd",
    "Fastflow",
    "Fre",
+    "Fuvas",


The __all__ list should maintain alphabetical order. Fuvas appears after Fre but Ganomaly follows immediately after. The ordering is inconsistent with the alphabetical pattern. Move Fuvas after Fre but ensure all entries maintain strict alphabetical order.

Copilot · 2025-12-04T17:00:51Z

src/anomalib/models/components/dinov2/vision_transformer.py

+        w0 = w // self.patch_size
+        h0 = h // self.patch_size
+        m = int(math.sqrt(n_pos))
+        assert n_pos == m * m


The assertion lacks an error message. If the assertion fails, users will see only 'AssertionError' without context. Add a descriptive message: assert n_pos == m * m, f'Expected {m * m} positional embeddings but got {n_pos}'

Suggested change

assert n_pos == m * m

assert n_pos == m * m, f'Expected {m * m} positional embeddings but got {n_pos}'

Copilot · 2025-12-04T17:00:51Z

src/anomalib/models/components/dinov2/layers/patch_embed.py

+        A tuple ``(h, w)``.
+    """
+    if isinstance(x, tuple):
+        assert len(x) == 2


The assertion lacks an error message explaining why the tuple must have length 2. Add a descriptive message: assert len(x) == 2, f'Expected tuple of length 2, got {len(x)}'

Suggested change

assert len(x) == 2

assert len(x) == 2, f'Expected tuple of length 2, got {len(x)}'

Copilot · 2025-12-04T17:00:51Z

src/anomalib/models/components/dinov2/layers/dino_head.py

+    if nlayers == 1:
+        return nn.Linear(in_dim, bottleneck_dim, bias=bias)
+
+    assert hidden_dim is not None, "hidden_dim must be provided when nlayers > 1"


The error message could be more helpful by indicating the actual value of nlayers. Change to: assert hidden_dim is not None, f'hidden_dim must be provided when nlayers ({nlayers}) > 1'

Suggested change

assert hidden_dim is not None, "hidden_dim must be provided when nlayers > 1"

assert hidden_dim is not None, f"hidden_dim must be provided when nlayers ({nlayers}) > 1"

codecov-commenter · 2025-12-04T17:15:36Z

⚠️ Please install the to ensure uploads and comments are reliably processed by Codecov.

Codecov Report

❌ Patch coverage is 77.73050% with 157 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
.../anomalib/models/components/dinov2/layers/block.py	52.11%	34 Missing ⚠️
...lib/models/components/dinov2/vision_transformer.py	82.12%	32 Missing ⚠️
...malib/models/components/dinov2/layers/dino_head.py	24.32%	28 Missing ⚠️
...alib/models/components/dinov2/layers/swiglu_ffn.py	32.43%	25 Missing ⚠️
...malib/models/components/dinov2/layers/drop_path.py	35.29%	11 Missing ⚠️
...lib/models/components/dinov2/layers/patch_embed.py	75.55%	11 Missing ⚠️
...malib/models/components/dinov2/layers/attention.py	80.95%	8 Missing ⚠️
.../anomalib/models/image/anomaly_dino/torch_model.py	93.93%	6 Missing ⚠️
...malib/models/image/anomaly_dino/lightning_model.py	96.07%	2 Missing ⚠️

📢 Thoughts on this report? Let us know!

Signed-off-by: Barabanov, Alexander <[email protected]>

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

Signed-off-by: Barabanov, Alexander <[email protected]>

Copilot

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

alfieroddanintel and others added 2 commits December 4, 2025 16:50

fix(dinov2_loader): update semgrep skip

9c36172

Copilot AI review requested due to automatic review settings December 4, 2025 16:44

rajeshgangireddy requested review from ashwinvaidya17 and samet-akcay as code owners December 4, 2025 16:44

Copilot AI reviewed Dec 4, 2025

View reviewed changes

github-advanced-security bot found potential problems Dec 4, 2025

View reviewed changes

fix(dinov2_loader): fixes comments

cdacd08

github-advanced-security bot found potential problems Dec 4, 2025

View reviewed changes

src/anomalib/models/components/dinov2/dinov2_loader.py Fixed Show fixed Hide fixed

fix(dinov2_loader): adjust comment placement for clarity in weight lo…

b35e3c4

…ading

Copilot AI review requested due to automatic review settings December 4, 2025 16:59

Copilot AI reviewed Dec 4, 2025

View reviewed changes

fix(model): resolve copilot comments

6715f54

AlexanderBarabanov mentioned this pull request Dec 5, 2025

semgrep ignores ignore comment open-edge-platform/geti-ci#26

Open

fix ci

16dfc5e

Signed-off-by: Barabanov, Alexander <[email protected]>

Copilot AI review requested due to automatic review settings December 5, 2025 22:40

AlexanderBarabanov requested a review from mramotowski as a code owner December 5, 2025 22:40

Copilot AI reviewed Dec 5, 2025

View reviewed changes

AlexanderBarabanov added 2 commits December 5, 2025 22:44

fix security checks on push

84aa738

Signed-off-by: Barabanov, Alexander <[email protected]>

suppress finding

3a5cb95

Signed-off-by: Barabanov, Alexander <[email protected]>

Copilot AI review requested due to automatic review settings December 5, 2025 22:56

Copilot AI reviewed Dec 5, 2025

View reviewed changes

	assert n_pos == m * m
	assert n_pos == m * m, f'Expected {m * m} positional embeddings but got {n_pos}'

	assert len(x) == 2
	assert len(x) == 2, f'Expected tuple of length 2, got {len(x)}'

	assert hidden_dim is not None, "hidden_dim must be provided when nlayers > 1"
	assert hidden_dim is not None, f"hidden_dim must be provided when nlayers ({nlayers}) > 1"

feat(anomalydino): Introduces anomalydino with semgrep "fixes" #3193

Are you sure you want to change the base?

feat(anomalydino): Introduces anomalydino with semgrep "fixes" #3193

Conversation

rajeshgangireddy commented Dec 4, 2025

📝 Description

✨ Changes

✅ Checklist

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Uh oh!

Copilot AI Dec 4, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 4, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 4, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 4, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 4, 2025

Choose a reason for hiding this comment

Uh oh!

codecov-commenter commented Dec 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

codecov-commenter commented Dec 4, 2025 •

edited

Loading