A library for benchmarking, developing and deploying deep learning anomaly detection algorithms

OpenVINO Toolkit

Last update: Jan 4, 2023

Related tags

Algorithms unsupervised-learning anomaly-detection neural-network-compression openvino anomaly-segmentation anomaly-localization

Overview

A library for benchmarking, developing and deploying deep learning anomaly detection algorithms

Key Features • Getting Started • Docs • License

Introduction

Anomalib is a deep learning library that aims to collect state-of-the-art anomaly detection algorithms for benchmarking on both public and private datasets. Anomalib provides several ready-to-use implementations of anomaly detection algorithms described in the recent literature, as well as a set of tools that facilitate the development and implementation of custom models. The library has a strong focus on image-based anomaly detection, where the goal of the algorithm is to identify anomalous images, or anomalous pixel regions within images in a dataset. Anomalib is constantly updated with new algorithms and training/inference extensions, so keep checking!

Key features:

The largest public collection of ready-to-use deep learning anomaly detection algorithms and benchmark datasets.
PyTorch Lightning based model implementations to reduce boilerplate code and limit the implementation efforts to the bare essentials.
All models can be exported to OpenVINO Intermediate Representation (IR) for accelerated inference on intel hardware.
A set of inference tools for quick and easy deployment of the standard or custom anomaly detection models.

Getting Started

To get an overview of all the devices where anomalib as been tested thoroughly, look at the Supported Hardware section in the documentation.

PyPI Install

You can get started with anomalib by just using pip.

pip install anomalib

Local Install

It is highly recommended to use virtual environment when installing anomalib. For instance, with anaconda, anomalib could be installed as,

yes | conda create -n anomalib_env python=3.8
conda activate anomalib_env
git clone https://github.com/openvinotoolkit/anomalib.git
cd anomalib
pip install -e .

Training

By default python tools/train.py runs PADIM model MVTec leather dataset.

python tools/train.py    # Train PADIM on MVTec leather

Training a model on a specific dataset and category requires further configuration. Each model has its own configuration file, config.yaml , which contains data, model and training configurable parameters. To train a specific model on a specific dataset and category, the config file is to be provided:

python tools/train.py --model_config_path <path/to/model/config.yaml>

For example, to train PADIM you can use

python tools/train.py --model_config_path anomalib/models/padim/config.yaml

Alternatively, a model name could also be provided as an argument, where the scripts automatically finds the corresponding config file.

python tools/train.py --model padim

where the currently available models are:

Inference

Anomalib contains several tools that can be used to perform inference with a trained model. The script in tools/inference contains an example of how the inference tools can be used to generate a prediction for an input image.

If the specified weight path points to a PyTorch Lightning checkpoint file (.ckpt), inference will run in PyTorch. If the path points to an ONNX graph (.onnx) or OpenVINO IR (.bin or .xml), inference will run in OpenVINO.

The following command can be used to run inference from the command line:

python tools/inference.py \
    --model_config_path <path/to/model/config.yaml> \
    --weight_path <path/to/weight/file> \
    --image_path <path/to/image>

As a quick example:

python tools/inference.py \
    --model_config_path anomalib/models/padim/config.yaml \
    --weight_path results/padim/mvtec/bottle/weights/model.ckpt \
    --image_path datasets/MVTec/bottle/test/broken_large/000.png

If you want to run OpenVINO model, ensure that compression apply is set to True in the respective model config.yaml.

optimization:
  compression:
    apply: true

Example OpenVINO Inference:

python tools/inference.py \
    --model_config_path  \
    anomalib/models/padim/config.yaml  \
    --weight_path  \
    results/padim/mvtec/bottle/compressed/compressed_model.xml  \
    --image_path  \
    datasets/MVTec/bottle/test/broken_large/000.png  \
    --meta_data  \
    results/padim/mvtec/bottle/compressed/meta_data.json

Ensure that you provide path to meta_data.json if you want the normalization to be applied correctly.

Datasets

MVTec Dataset

Image-Level AUC

Model		Avg	Carpet	Grid	Leather	Tile	Wood	Bottle	Cable	Capsule	Hazelnut	Metal Nut	Pill	Screw	Toothbrush	Transistor	Zipper
PatchCore	Wide ResNet-50	0.980	0.984	0.959	1.000	1.000	0.989	1.000	0.990	0.982	1.000	0.994	0.924	0.960	0.933	1.000	0.982
PatchCore	ResNet-18	0.973	0.970	0.947	1.000	0.997	0.997	1.000	0.986	0.965	1.000	0.991	0.916	0.943	0.931	0.996	0.953
CFlow	Wide ResNet-50	0.962	0.986	0.962	1.0	0.999	0.993	1.0	0.893	0.945	1.0	0.995	0.924	0.908	0.897	0.943	0.984
PaDiM	Wide ResNet-50	0.950	0.995	0.942	1.0	0.974	0.993	0.999	0.878	0.927	0.964	0.989	0.939	0.845	0.942	0.976	0.882
PaDiM	ResNet-18	0.891	0.945	0.857	0.982	0.950	0.976	0.994	0.844	0.901	0.750	0.961	0.863	0.759	0.889	0.920	0.780
STFPM	Wide ResNet-50	0.876	0.957	0.977	0.981	0.976	0.939	0.987	0.878	0.732	0.995	0.973	0.652	0.825	0.5	0.875	0.899
STFPM	ResNet-18	0.893	0.954	0.982	0.989	0.949	0.961	0.979	0.838	0.759	0.999	0.956	0.705	0.835	0.997	0.853	0.645
DFM	Wide ResNet-50	0.891	0.978	0.540	0.979	0.977	0.974	0.990	0.891	0.931	0.947	0.839	0.809	0.700	0.911	0.915	0.981
DFM	ResNet-18	0.894	0.864	0.558	0.945	0.984	0.946	0.994	0.913	0.871	0.979	0.941	0.838	0.761	0.95	0.911	0.949
DFKDE	Wide ResNet-50	0.774	0.708	0.422	0.905	0.959	0.903	0.936	0.746	0.853	0.736	0.687	0.749	0.574	0.697	0.843	0.892
DFKDE	ResNet-18	0.762	0.646	0.577	0.669	0.965	0.863	0.951	0.751	0.698	0.806	0.729	0.607	0.694	0.767	0.839	0.866

Pixel-Level AUC

Model		Avg	Carpet	Grid	Leather	Tile	Wood	Bottle	Cable	Capsule	Hazelnut	Metal Nut	Pill	Screw	Toothbrush	Transistor	Zipper
PatchCore	Wide ResNet-50	0.980	0.988	0.968	0.991	0.961	0.934	0.984	0.988	0.988	0.987	0.989	0.980	0.989	0.988	0.981	0.983
PatchCore	ResNet-18	0.976	0.986	0.955	0.990	0.943	0.933	0.981	0.984	0.986	0.986	0.986	0.974	0.991	0.988	0.974	0.983
CFlow	Wide ResNet-50	0.971	0.986	0.968	0.993	0.968	0.924	0.981	0.955	0.988	0.990	0.982	0.983	0.979	0.985	0.897	0.980
PaDiM	Wide ResNet-50	0.979	0.991	0.970	0.993	0.955	0.957	0.985	0.970	0.988	0.985	0.982	0.966	0.988	0.991	0.976	0.986
PaDiM	ResNet-18	0.968	0.984	0.918	0.994	0.934	0.947	0.983	0.965	0.984	0.978	0.970	0.957	0.978	0.988	0.968	0.979
STFPM	Wide ResNet-50	0.903	0.987	0.989	0.980	0.966	0.956	0.966	0.913	0.956	0.974	0.961	0.946	0.988	0.178	0.807	0.980
STFPM	ResNet-18	0.951	0.986	0.988	0.991	0.946	0.949	0.971	0.898	0.962	0.981	0.942	0.878	0.983	0.983	0.838	0.972

Image F1 Score

Model		Avg	Carpet	Grid	Leather	Tile	Wood	Bottle	Cable	Capsule	Hazelnut	Metal Nut	Pill	Screw	Toothbrush	Transistor	Zipper
PatchCore	Wide ResNet-50	0.976	0.971	0.974	1.000	1.000	0.967	1.000	0.968	0.982	1.000	0.984	0.940	0.943	0.938	1.000	0.979
PatchCore	ResNet-18	0.970	0.949	0.946	1.000	0.98	0.992	1.000	0.978	0.969	1.000	0.989	0.940	0.932	0.935	0.974	0.967
CFlow	Wide ResNet-50	0.944	0.972	0.932	1.0	0.988	0.967	1.0	0.832	0.939	1.0	0.979	0.924	0.971	0.870	0.818	0.967
PaDiM	Wide ResNet-50	0.951	0.989	0.930	1.0	0.960	0.983	0.992	0.856	0.982	0.937	0.978	0.946	0.895	0.952	0.914	0.947
PaDiM	ResNet-18	0.916	0.930	0.893	0.984	0.934	0.952	0.976	0.858	0.960	0.836	0.974	0.932	0.879	0.923	0.796	0.915
STFPM	Wide ResNet-50	0.926	0.973	0.973	0.974	0.965	0.929	0.976	0.853	0.920	0.972	0.974	0.922	0.884	0.833	0.815	0.931
STFPM	ResNet-18	0.932	0.961	0.982	0.989	0.930	0.951	0.984	0.819	0.918	0.993	0.973	0.918	0.887	0.984	0.790	0.908
DFM	Wide ResNet-50	0.918	0.960	0.844	0.990	0.970	0.959	0.976	0.848	0.944	0.913	0.912	0.919	0.859	0.893	0.815	0.961
DFM	ResNet-18	0.919	0.895	0.844	0.926	0.971	0.948	0.977	0.874	0.935	0.957	0.958	0.921	0.874	0.933	0.833	0.943
DFKDE	Wide ResNet-50	0.875	0.907	0.844	0.905	0.945	0.914	0.946	0.790	0.914	0.817	0.894	0.922	0.855	0.845	0.722	0.910
DFKDE	ResNet-18	0.872	0.864	0.844	0.854	0.960	0.898	0.942	0.793	0.908	0.827	0.894	0.916	0.859	0.853	0.756	0.916

Comments

1 epoch

Is it meant to be only one epoch in training? Your config files state 1 epoch, is that just as quick example? I tried to train PADIM for 10 on mvtec leather and wood and the metrics stay the same anyway, so it seems nothing as gained by training more. Lightning module also warn that there is no optimizer so I guess train only finds the correct thresholds and that takes 1 epoch.

opened by sequoiagrove 20
Improving the result for custom dataset

Hi, I am able to running the code but when I am working on the custom dataset, the code did not achive to detect the defect. I tried change the threshold values ,epoch numbers etc.. Do you have any advice for getting good result on my custom dataset? Thanks,
Metrics

opened by ZeynepRuveyda 18
add option to load metrics with kwargs
Description

Fixes #687

Changes

[ ] Bug fix (non-breaking change which fixes an issue)

[ ] Refactor (non-breaking change which refactors the code base)

[X] New feature (non-breaking change which adds functionality)

[ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)

[X] This change requires a documentation update

Checklist

[X] My code follows the pre-commit style and check guidelines of this project.

[X] I have performed a self-review of my code

[X] I have commented my code, particularly in hard-to-understand areas

[ ] I have made corresponding changes to the documentation

[ ] My changes generate no new warnings

[x] I have added tests that prove my fix is effective or that my feature works

[x] New and existing tests pass locally with my changes

Metrics Callbacks Tests Docs
opened by jpcbertoldo 15
Fix #699
Fixes #699

Changes

[X] Bug fix (non-breaking change which fixes an issue)

[ ] Refactor (non-breaking change which refactors the code base)

[ ] New feature (non-breaking change which adds functionality)

[ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)

[ ] This change requires a documentation update

Checklist

[] My code follows the pre-commit style and check guidelines of this project.

[ ] I have performed a self-review of my code

[ ] I have commented my code, particularly in hard-to-understand areas

[ ] I have made corresponding changes to the documentation

[ ] My changes generate no new warnings

[ ] I have added tests that prove my fix is effective or that my feature works

[ ] New and existing tests pass locally with my changes
opened by jpcbertoldo 14

TypeError: unsupported operand type(s) for /: 'NoneType' and 'float'

I want to train CFLOW model on custom dataset.

config.yaml

dataset:
  name: Concrete_Crack #options: [mvtec, btech, folder, concrete_crack]
  format: folder # mvtec
  path: ./datasets/Concrete_Crack/ # ./datasets/MVTec
  normal_dir: 'train/Negative'
  abnormal_dir: 'test/Positive'
  normal_test_dir: 'test/Negative'
  task: segmentation
  mask: ./datasets/Concrete_Crack/ground_truth/
  extensions: '.jpg'
  split_ratio: 0.1
  seed: 0
#  category: bottle
  image_size: 227
  train_batch_size: 8 # 16
  test_batch_size: 8 # 16
  inference_batch_size: 8 # 16
  fiber_batch_size: 64
  num_workers: 8
  transform_config:
    train: null
    val: null
  create_validation_set: false

My dataset have the same structure like MVTec

To Reproduce python3 tools/train.py --model_config_path anomalib/models/cflow/config.yaml

When I start the training, I get an error:

File "/home/Projects/Anomalib/anomalib/anomalib/data/folder.py", line 282, in __getitem__
    mask = cv2.imread(mask_path, flags=0) / 255.0
TypeError: unsupported operand type(s) for /: 'NoneType' and 'float'

Data

opened by andriy-onufriyenko 14

PatchCore results are much worth than reported
Describe the bug

A clear and concise description of what the bug is.

To Reproduce

Steps to reproduce the behavior:

Go to the Main directory

Run python tools/train.py --model patchcore

Expected behavior

The image AUCROC to be .98 for the category carpet of the MvTech dataset but it is very low. Fastflow work as expected so the problem seems to be patchcore. Screenshots

Hardware and Software Configuration

OS: [Ubuntu]

NVIDIA Driver Version [470.141.03]

CUDA Version [11.4]

CUDNN Version [e.g. v11.4.120]

Log

WARNING: CPU random generator seem to be failing, disabling hardware random number generation WARNING: RDRND generated: 0xffffffff 0xffffffff 0xffffffff 0xffffffff ----------------------------------/anomalib/config/config.py:166: UserWarning: config.project.unique_dir is set to False. This does not ensure that your results will be written in an empty directory and you may overwrite files. warn( 2022-11-16 11:52:49,662 - anomalib.data - INFO - Loading the datamodule 2022-11-16 11:52:49,662 - anomalib.pre_processing.pre_process - WARNING - Transform configs has not been provided. Images will be normalized using ImageNet statistics. 2022-11-16 11:52:49,663 - anomalib.pre_processing.pre_process - WARNING - Transform configs has not been provided. Images will be normalized using ImageNet statistics. 2022-11-16 11:52:49,663 - anomalib.models - INFO - Loading the model. 2022-11-16 11:52:49,667 - torch.distributed.nn.jit.instantiator - INFO - Created a temporary directory at /tmp/tmpk5fh8j6r 2022-11-16 11:52:49,667 - torch.distributed.nn.jit.instantiator - INFO - Writing /tmp/tmpk5fh8j6r/_remote_module_non_scriptable.py 2022-11-16 11:52:49,674 - anomalib.models.components.base.anomaly_module - INFO - Initializing PatchcoreLightning model. /home/-/code/anomalib/lib/python3.8/site-packages/torchmetrics/utilities/prints.py:36: UserWarning: Metric PrecisionRecallCurve will save all targets and predictions in buffer. For large datasets this may lead to large memory footprint. warnings.warn(*args, **kwargs) 2022-11-16 11:52:50,882 - timm.models.helpers - INFO - Loading pretrained weights from url (https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/wide_resnet50_racm-8234f177.pth) 2022-11-16 11:52:51,009 - anomalib.utils.loggers - INFO - Loading the experiment logger(s) 2022-11-16 11:52:51,009 - anomalib.utils.callbacks - INFO - Loading the callbacks /home/-/code/anomalib/src/anomalib/anomalib/utils/callbacks/init.py:141: UserWarning: Export option: None not found. Defaulting to no model export warnings.warn(f"Export option: {config.optimization.export_mode} not found. Defaulting to no model export") 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - GPU available: True, used: True 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - TPU available: False, using: 0 TPU cores 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - IPU available: False, using: 0 IPUs 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - HPU available: False, using: 0 HPUs 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - Trainer(limit_train_batches=1.0) was configured so 100% of the batches per epoch will be used.. 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - Trainer(limit_val_batches=1.0) was configured so 100% of the batches will be used.. 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - Trainer(limit_test_batches=1.0) was configured so 100% of the batches will be used.. 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - Trainer(limit_predict_batches=1.0) was configured so 100% of the batches will be used.. 2022-11-16 11:52:51,012 - pytorch_lightning.utilities.rank_zero - INFO - Trainer(val_check_interval=1.0) was configured so validation will run at the end of the training epoch.. 2022-11-16 11:52:51,012 - anomalib - INFO - Training the model. 2022-11-16 11:52:51,016 - anomalib.data.mvtec - INFO - Found the dataset. 2022-11-16 11:52:51,018 - anomalib.data.mvtec - INFO - Setting up train, validation, test and prediction datasets. /-/-/code/anomalib/lib/python3.8/site-packages/torchmetrics/utilities/prints.py:36: UserWarning: Metric ROC will save all targets and predictions in buffer. For large datasets this may lead to large memory footprint. warnings.warn(*args, **kwargs) 2022-11-16 11:52:52,479 - pytorch_lightning.accelerators.gpu - INFO - LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1] /-/-/code/anomalib/lib/python3.8/site-packages/pytorch_lightning/core/optimizer.py:183: UserWarning: LightningModule.configure_optimizers returned None, this fit will run with no optimizer rank_zero_warn( 2022-11-16 11:52:52,482 - pytorch_lightning.callbacks.model_summary - INFO - | Name | Type | Params

0 | image_threshold | AnomalyScoreThreshold | 0
1 | pixel_threshold | AnomalyScoreThreshold | 0
2 | model | PatchcoreModel | 24.9 M 3 | image_metrics | AnomalibMetricCollection | 0
4 | pixel_metrics | AnomalibMetricCollection | 0
5 | normalization_metrics | MinMax | 0

24.9 M Trainable params 0 Non-trainable params 24.9 M Total params 99.450 Total estimated model params size (MB) Epoch 0: 8%|▊ | 1/13 [00:01<00:16, 1.37s/it, loss=nan]/-/-/code/anomalib/lib/python3.8/site-packages/pytorch_lightning/loops/optimization/optimizer_loop.py:137: UserWarning: training_step returned None. If this was on purpose, ignore this warning... self.warning_cache.warn("training_step returned None. If this was on purpose, ignore this warning...") Epoch 0: 69%|██████▉ | 9/13 [00:01<00:00, 4.67it/s, loss=nan] Validation: 0it [00:00, ?it/s]2022-11-16 11:52:54,414 - anomalib.models.patchcore.lightning_model - INFO - Aggregating the embedding extracted from the training set. 2022-11-16 11:52:54,415 - anomalib.models.patchcore.lightning_model - INFO - Applying core-set subsampling to get the embedding. Epoch 0: 69%|██████▉ | 9/13 [00:20<00:08, 2.22s/it, loss=nan] Validation: 0%| | 0/4 [00:00<?, ?it/s] Validation DataLoader 0: 0%| | 0/4 [00:00<?, ?it/s] Validation DataLoader 0: 25%|██▌ | 1/4 [00:00<00:00, 4.13it/s] Epoch 0: 77%|███████▋ | 10/13 [00:59<00:17, 5.94s/it, loss=nan] Validation DataLoader 0: 50%|█████ | 2/4 [00:00<00:00, 4.04it/s] Epoch 0: 85%|████████▍ | 11/13 [00:59<00:10, 5.42s/it, loss=nan] Validation DataLoader 0: 75%|███████▌ | 3/4 [00:00<00:00, 4.02it/s] Epoch 0: 92%|█████████▏| 12/13 [00:59<00:04, 4.99s/it, loss=nan] Validation DataLoader 0: 100%|██████████| 4/4 [00:00<00:00, 4.63it/s] Epoch 0: 100%|██████████| 13/13 [01:00<00:00, 4.67s/it, loss=nan, pixel_F1Score=0.548, pixel_AUROC=0.986] Epoch 0: 100%|██████████| 13/13 [01:01<00:00, 4.69s/it, loss=nan, pixel_F1Score=0.548, pixel_AUROC=0.986] 2022-11-16 11:53:53,628 - anomalib.utils.callbacks.timer - INFO - Training took 61.15 seconds 2022-11-16 11:53:53,628 - anomalib - INFO - Loading the best model weights. 2022-11-16 11:53:53,628 - anomalib - INFO - Testing the model. 2022-11-16 11:53:53,632 - anomalib.data.mvtec - INFO - Found the dataset. 2022-11-16 11:53:53,633 - anomalib.data.mvtec - INFO - Setting up train, validation, test and prediction datasets. /-/code/anomalib/lib/python3.8/site-packages/torchmetrics/utilities/prints.py:36: UserWarning: Metric ROC will save all targets and predictions in buffer. For large datasets this may lead to large memory footprint. warnings.warn(*args, **kwargs) 2022-11-16 11:53:53,716 - pytorch_lightning.accelerators.gpu - INFO - LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1] 2022-11-16 11:53:53,718 - anomalib.utils.callbacks.model_loader - INFO - Loading the model from /home/-/code/anomalib/src/anomalib/results/patchcore/mvtec/carpet/run/weights/model.ckpt Testing DataLoader 0: 100%|██████████| 4/4 [00:19<00:00, 4.65s/it]2022-11-16 11:54:14,762 - anomalib.utils.callbacks.timer - INFO - Testing took 20.9255051612854 seconds Throughput (batch_size=32) : 5.591262867883519 FPS Testing DataLoader 0: 100%|██████████| 4/4 [00:19<00:00, 4.97s/it] ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── Test metric DataLoader 0 ──────────────────────────────────────────────────────────────────────────────────────────────────────────────────────── image_AUROC 0.4036917984485626 image_F1Score 0.8640776872634888 pixel_AUROC 0.9860672950744629 pixel_F1Score 0.5481611490249634 ────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

Process finished with exit code 0
Bug
opened by spoorgholi74 13
Migrate old configs to new CLI format
Description

Migrate old configs to match the new CLI design

TODO

[x] Compare performance after migration (ensure that all the optimizers and callbacks are loaded correctly)

[x] Remove LightningModels (they exist for backward compatibility) with the new design we can use the same one as LightningCLI

[x] Extend to cflow, dfkde, dfm, draem, fastflow, ganomaly, strpm, patchcore, reverse_distillation

[x] Modify tests

Callbacks Data Logger Config CLI Inference Tests Docs Benchmarking HPO Tools Notebooks
opened by ashwinvaidya17 13

Error(s) in loading state_dict for PatchcoreLightning

Describe the bug

Running tools/inference/torch_inference.py throws:

Traceback (most recent call last):
  File "/workspace/tools/inference/torch_inference.py", line 97, in <module>
    infer()
  File "/workspace/tools/inference/torch_inference.py", line 73, in infer
    inferencer = TorchInferencer(config=args.config, model_source=args.weights)
  File "/workspace/anomalib/deploy/inferencers/torch_inferencer.py", line 54, in __init__
    self.model = self.load_model(model_source)
  File "/workspace/anomalib/deploy/inferencers/torch_inferencer.py", line 85, in load_model
    model.load_state_dict(torch.load(path)["state_dict"])
  File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1497, in load_state_dict
    raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format(
RuntimeError: Error(s) in loading state_dict for PatchcoreLightning:
	Unexpected key(s) in state_dict: "normalization_metrics.min", "normalization_metrics.max".

To Reproduce

Using default patchcore/config.yaml
Train the model: python tools/train.py --config anomalib/models/patchcore/config.yaml
Use trained model for torch_inference:

python tools/inference/torch_inference.py \
--config anomalib/models/patchcore/config.yaml \
--weights results/patchcore/mvtec/bottle/weights/model-v14.ckpt \
--input datasets/MVTec/bottle/test/broken_large/000.png \
--output results/patchcore/mvtec/bottle/images

Additional context

Error is present since commit d6951ebaaf36477bf245b17b5b6a563d35892e81

opened by Luksalos 13

Issues with heatmap scaling when training without masks

I ran into an issue when training on a dataset (subset of MVTec), where I set all ground truth masks to zero (to simulate training on a dataset for which I have no ground truth masks). When training with the actual ground truth masks, the model produces heatmaps as expected as in the first image below (produced with tools/inference.py). However, when training with the zero masks, the heatmaps seem to be scaled differently as in the second image below. The confidence score seems unaffected.

ground_truth_masks

zero_masks

This behaviour is the same for both PADIM and PatchCore. I haven't tested the other models.

This is my model config for PADIM

dataset:
  name: mvtec_test
  format: folder
  path: ./datasets/mvtec_test/images
  normal: normal
  abnormal: abnormal
  task: segmentation
  mask: ./datasets/mvtec_test/masks_orig
  extensions: null
  seed: 0  
  image_size: 224
  train_batch_size: 32
  test_batch_size: 1
  num_workers: 16
  transform_config: null
  split_ratio: 0.2
  create_validation_set: true
  tiling:
    apply: false
    tile_size: null
    stride: null
    remove_border_count: 0
    use_random_tiling: False
    random_tile_count: 16

model:
  name: padim
  backbone: resnet18
  layers:
    - layer1
    - layer2
    - layer3
  metric: auc
  normalization_method: min_max # options: [none, min_max, cdf]
  threshold:
    image_default: 3
    pixel_default: 3
    adaptive: true

project:
  seed: 42
  path: ./results
  log_images_to: ["local"]
  logger: false
  save_to_csv: false

optimization:
  openvino:
    apply: false

# PL Trainer Args. Don't add extra parameter here.
trainer:
  accelerator: null
  accumulate_grad_batches: 1
  amp_backend: native
  auto_lr_find: false
  auto_scale_batch_size: false
  auto_select_gpus: false
  benchmark: false
  check_val_every_n_epoch: 1 # Don't validate before extracting features.
  checkpoint_callback: true
  default_root_dir: null
  deterministic: false
  fast_dev_run: false
  gpus: 1
  gradient_clip_val: 0
  limit_predict_batches: 1.0
  limit_test_batches: 1.0
  limit_train_batches: 1.0
  limit_val_batches: 1.0
  log_every_n_steps: 50
  log_gpu_memory: null
  max_epochs: 1
  max_steps: -1
  min_epochs: null
  min_steps: null
  move_metrics_to_cpu: false
  multiple_trainloader_mode: max_size_cycle
  num_nodes: 1
  num_processes: 1
  num_sanity_val_steps: 0
  overfit_batches: 0.0
  plugins: null
  precision: 32
  prepare_data_per_node: true
  process_position: 0
  profiler: null
  progress_bar_refresh_rate: null
  replace_sampler_ddp: true
  stochastic_weight_avg: false
  sync_batchnorm: false
  terminate_on_nan: false
  tpu_cores: null
  track_grad_norm: -1
  val_check_interval: 1.0 # Don't validate before extracting features.
  weights_save_path: null
  weights_summary: top

opened by LukasBommes 13

Support Kaggle and Colab Environments
Is your feature request related to a problem? Please describe.

Kaggle and Colab are free and great resources to do quick experiments. Currently, I've found that it's not possible to run anomalib on these environments. I guess, when the PIP packages get ready, it would be possible. Supporting these environments would be a great opportunity for the end-user for sure.
opened by innat 13
How to train in a custom dataset?

I am having a bit of trouble training on a custom dataset. I reproduced the MVTec structure but I still get some errors like:

IndexError: too many indices for tensor of dimension 0

opened by opassos 13
CVE-2007-4559 Patch

Patching CVE-2007-4559

Hi, we are security researchers from the Advanced Research Center at Trellix. We have began a campaign to patch a widespread bug named CVE-2007-4559. CVE-2007-4559 is a 15 year old bug in the Python tarfile package. By using extract() or extractall() on a tarfile object without sanitizing input, a maliciously crafted .tar file could perform a directory path traversal attack. We found at least one unsantized extractall() in your codebase and are providing a patch for you via pull request. The patch essentially checks to see if all tarfile members will be extracted safely and throws an exception otherwise. We encourage you to use this patch or your own solution to secure against CVE-2007-4559. Further technical information about the vulnerability can be found in this blog.

If you have further questions you may contact us through this projects lead researcher Kasimir Schulz.
Data

opened by TrellixVulnTeam 0
[Detection] Compute box score when generating boxes from masks
Description

Small PR that adds an anomaly score for bounding boxes generated from anomaly maps, which is needed for the OTX task.

Extends masks_to_boxes to take anomaly maps as optional additional input, and compute an anomaly score for each bounding box as the max value of the anomaly map within the pixel region covered by the bounding box.

Update the logic in AnomalyModule.

Changes

[ ] Bug fix (non-breaking change which fixes an issue)

[ ] Refactor (non-breaking change which refactors the code base)

[x] New feature (non-breaking change which adds functionality)

[ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)

[ ] This change requires a documentation update

Checklist

[x] My code follows the pre-commit style and check guidelines of this project.

[x] I have performed a self-review of my code

[ ] I have commented my code, particularly in hard-to-understand areas

[ ] I have made corresponding changes to the documentation

[x] My changes generate no new warnings

[x] I have added tests that prove my fix is effective or that my feature works

[x] New and existing tests pass locally with my changes

[ ] I have added a summary of my changes to the CHANGELOG (not for minor changes, docs and tests).

Data Inference Tests
opened by djdameln 0
[WIP] CLI refactor
Description

In progress marked as draft as this contains changes from https://github.com/openvinotoolkit/anomalib/pull/713 and is targeting that branch for now.

Done

[x] Add HPO and Benchmarking to the CLI

[x] Migrate old configs to the new format

[x] Basic train,test,hpo etc work with the new CLI

TODO

[ ] Expose internal configuration in benchmarking and hpo scripts.

[ ] Add tests for CLI

[ ] Test logging from CLI

[ ] Test if all notebooks run with these changes

[ ] Use the new @tiler decorator in the models

[ ] Refactor code (especially the new callback stuff)

[ ] Potentially investigate replacing lightning cli with custom one based on json argparse

[ ] Add custom loops

Changes

[ ] Bug fix (non-breaking change which fixes an issue)

[ ] Refactor (non-breaking change which refactors the code base)

[ ] New feature (non-breaking change which adds functionality)

[ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)

[ ] This change requires a documentation update

Checklist

[ ] My code follows the pre-commit style and check guidelines of this project.

[ ] I have performed a self-review of my code

[ ] I have commented my code, particularly in hard-to-understand areas

[ ] I have made corresponding changes to the documentation

[ ] My changes generate no new warnings

[ ] I have added tests that prove my fix is effective or that my feature works

[ ] New and existing tests pass locally with my changes

[ ] I have added a summary of my changes to the CHANGELOG (not for minor changes, docs and tests).

Callbacks Pre-Processing Config CLI Inference Tests Benchmarking HPO Tools
opened by ashwinvaidya17 0

[Bug]: how to use tiler in training / inference

Describe the bug

We try to enable tiling in the training / inference, by setting configs like this:

  tiling:
    apply: true #false
    tile_size: 224 #null

however, setting this doesn't actually enable the tiler. Closely reading the code, we've seen anomalib/models/patchcore/torch_model.py actually tiler is not set at all

    def __init__(
        self,
        input_size: Tuple[int, int],
        layers: List[str],
        backbone: str = "wide_resnet50_2",
        pre_trained: bool = True,
        num_neighbors: int = 9,
    ) -> None:
        super().__init__()
        self.tiler: Optional[Tiler] = None   ######### here is None

        self.backbone = backbone
        self.layers = layers
        self.input_size = input_size
        self.num_neighbors = num_neighbors

could you please suggest how to enable tiling? Thanks!

Dataset

MVTec

Model

PatchCore

Steps to reproduce the behavior

N/A

OS information

OS information:

OS: [e.g. Ubuntu 20.04]
Python version: [e.g. 3.8.10]
Anomalib version: [e.g. 0.3.6]
PyTorch version: [e.g. 1.9.0]
CUDA/cuDNN version: [e.g. 11.1]
GPU models and configuration: [e.g. 2x GeForce RTX 3090]
Any other relevant information: [e.g. I'm using a custom dataset] N/A

Expected behavior

Enable tiling in train/test/inference.

Screenshots

No response

Pip/GitHub

pip

What version/branch did you use?

No response

Configuration YAML

dataset:
  name: mvtec #options: [mvtec, btech, folder]
  format: mvtec
  path: ./datasets/MVTec
  task: segmentation
  category: bottle   
  image_size: 336 #224
  train_batch_size: 2 #32
  test_batch_size: 2 #32
  num_workers: 8
  transform_config:
    train: null
    val: null
  create_validation_set: false
  tiling:
    apply: true #false
    tile_size: 224 #null
    stride: null
    remove_border_count: 0
    use_random_tiling: False
    random_tile_count: 16

model:
  name: patchcore
  backbone: wide_resnet50_2
  pre_trained: true
  layers:
    - layer2
    - layer3
  coreset_sampling_ratio: 0.1
  num_neighbors: 9
  normalization_method: min_max # options: [null, min_max, cdf]

metrics:
  image:
    - F1Score
    - AUROC
  pixel:
    - F1Score
    - AUROC
  threshold:
    #method: adaptive #options: [adaptive, manual]
    method: manual
    manual_image: 0.2 #null
    manual_pixel: 0.2 # null

visualization:
  show_images: False # show images on the screen
  save_images: True # save images to the file system
  log_images: True # log images to the available loggers (if any)
  image_save_path: null # path to which images will be saved
  mode: full # options: ["full", "simple"]

project:
  seed: 0
  path: ./results

logging:
  logger: [] # options: [comet, tensorboard, wandb, csv] or combinations.
  log_graph: false # Logs the model graph to respective logger.

optimization:
  export_mode: null # options: onnx, openvino

# PL Trainer Args. Don't add extra parameter here.
trainer:
  accelerator: auto # <"cpu", "gpu", "tpu", "ipu", "hpu", "auto">
  accumulate_grad_batches: 1
  amp_backend: native
  auto_lr_find: false
  auto_scale_batch_size: false
  auto_select_gpus: false
  benchmark: false
  check_val_every_n_epoch: 1 # Don't validate before extracting features.
  default_root_dir: null
  detect_anomaly: false
  deterministic: false
  devices: 1
  enable_checkpointing: true
  enable_model_summary: true
  enable_progress_bar: true
  fast_dev_run: false
  gpus: null # Set automatically
  gradient_clip_val: 0
  ipus: null
  limit_predict_batches: 1.0
  limit_test_batches: 1.0
  limit_train_batches: 1.0
  limit_val_batches: 1.0
  log_every_n_steps: 50
  log_gpu_memory: null
  max_epochs: 1
  max_steps: -1
  max_time: null
  min_epochs: null
  min_steps: null
  move_metrics_to_cpu: false
  multiple_trainloader_mode: max_size_cycle
  num_nodes: 1
  num_processes: null
  num_sanity_val_steps: 0
  overfit_batches: 0.0
  plugins: null
  precision: 32
  profiler: null
  reload_dataloaders_every_n_epochs: 0
  replace_sampler_ddp: true
  strategy: null
  sync_batchnorm: false
  tpu_cores: null
  track_grad_norm: -1
  val_check_interval: 1.0 # Don't validate before extracting features.

Logs

N/A

Code of Conduct

[X] I agree to follow this project's Code of Conduct

opened by frankmanbb 0

[Dataset] Add VisA dataset
Description

This PR adds the Visual Anomaly (VisA) dataset.

The dataset follows the same format as MVTec, so we could re-use the make_mvtec_dataset function.

The make_mvtec_dataset function was slightly modified to make the mask file naming convention a bit more flexible (mvtec uses "000_mask.png", while visa uses "000.png").

There was a lot of duplication in the download and extract functionality of the different datasets, so this was moved to a shared location.

Currently targeted to feature branch, but will re-target to main once #822 has been merged.

Some examples:

Known Issues

~~CI will probably fail, because the dataset is not yet installed on the CI machine.~~

Changes

[ ] Bug fix (non-breaking change which fixes an issue)

[ ] Refactor (non-breaking change which refactors the code base)

[x] New feature (non-breaking change which adds functionality)

[ ] Breaking change (fix or feature that would cause existing functionality to not work as expected)

[ ] This change requires a documentation update

Checklist

[x] My code follows the pre-commit style and check guidelines of this project.

[x] I have performed a self-review of my code

[x] I have commented my code, particularly in hard-to-understand areas

[ ] I have made corresponding changes to the documentation

[x] My changes generate no new warnings

[x] I have added tests that prove my fix is effective or that my feature works

[x] New and existing tests pass locally with my changes

[x] I have added a summary of my changes to the CHANGELOG (not for minor changes, docs and tests).

Data Config Tests Docs
opened by djdameln 0

Releases(rkde-weights)

rkde-weights(Jan 2, 2023)

This release adds the feature extractor weights for the RKDE model implementation.
Source code(tar.gz)
Source code(zip)
rkde_feature_extractor.pth(217.45 MB)
v0.3.7(Oct 28, 2022)
What's Changed

Feature/comet logging by @sherpan in https://github.com/openvinotoolkit/anomalib/pull/517

🐞 Fix linting issues by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/535

🐞 Bug Fix: Solve NaN values of anomaly scores for PatchCore model by @bsl546 in https://github.com/openvinotoolkit/anomalib/pull/549

🐞 Bug Fix: Help description for argument task by @youngquan in https://github.com/openvinotoolkit/anomalib/pull/547

reutrn results of load_state_dict func by @zywvvd in https://github.com/openvinotoolkit/anomalib/pull/546

🔨 Pass pre-trained from config to ModelLightning by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/529

Benchmarking tool with Comet by @sherpan in https://github.com/openvinotoolkit/anomalib/pull/545

Add map_location when loading the weights by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/562

Add patchcore to openvino export test + upgrade lightning by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/565

🐞 Fix category check for folder dataset in anomalib CLI by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/567

Refactor PreProcessor and fix Visualizer denormalization issue. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/570

🔨 Check for successful openvino conversion by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/571

Comet HPO by @sherpan in https://github.com/openvinotoolkit/anomalib/pull/563

Fix patchcore image-level score computation by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/580

Fix anomaly map computation in CFlow when batch size is 1. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/589

Documentation refactor by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/576

✨ Add notebook for hpo by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/592

🐞 Fix comet HPO by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/597

✨ Replace keys from benchmarking script by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/595

Update README.md by @Owaiskhan9654 in https://github.com/openvinotoolkit/anomalib/pull/623

🐳 Containerize CI by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/616

add deprecation warning to denormalize class by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/629

Anomalib CLI Improvements - Update metrics and create post_processing section in the config file by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/607

Convert adaptive_threshold to Enum in configs by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/637

Create meta_data.json with ONNX export as well as OpenVINO export by @calebmm in https://github.com/openvinotoolkit/anomalib/pull/636

🖌 refactor export callback by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/640

🐞 Address docs build by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/639

Optimized inference with onnx for patchcore. by @acai66 in https://github.com/openvinotoolkit/anomalib/pull/652

New Contributors

@sherpan made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/517

@bsl546 made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/549

@youngquan made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/547

@zywvvd made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/546

@Owaiskhan9654 made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/623

@calebmm made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/636

@acai66 made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/652

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.3.6...v0.3.7
Source code(tar.gz)
Source code(zip)
v0.3.6(Sep 2, 2022)
What's Changed

Add publish workflow + update references to main by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/480

Fix Dockerfile by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/478

Fix onnx export by rewriting GaussianBlur by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/476

DFKDE refactor to accept any layer name like other models by @ashishbdatta in https://github.com/openvinotoolkit/anomalib/pull/482

🐞 Log benchmarking results in sub folder by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/483

🐞 Fix Visualization keys in new CLI by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/487

fix Perlin augmenter for non divisible image sizes by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/490

📝 Update the license headers by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/491

change default parameter values for DRAEM by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/495

Add reset methods to metrics by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/488

Feature Extractor Refactor by @ashishbdatta in https://github.com/openvinotoolkit/anomalib/pull/451

Convert AnomalyMapGenerator to nn.Module by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/497

Add github pr labeler to automatically label PRs by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/498

Add coverage by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/499

🐞 Change if check by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/501

SSPCAB implementation by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/500

🛠 Refactor Normalization by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/496

Enable generic exporting of a trained model to ONNX or OpenVINO IR by @ashishbdatta in https://github.com/openvinotoolkit/anomalib/pull/509

Updated documentation to add examples for exporting model by @ashishbdatta in https://github.com/openvinotoolkit/anomalib/pull/515

Ignore pixel metrics in classification task by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/516

Update export documentation by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/521

FIX: PaDiM didn't use config.model.pre_trained. by @jingt2ch in https://github.com/openvinotoolkit/anomalib/pull/514

Reset adaptive threshold between epochs by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/527

Add PRO metric by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/508

Set full_state_update attribute in custom metrics by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/531

🐞 Set normalization method from anomaly module by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/530

New Contributors

@ashishbdatta made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/482

@jingt2ch made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/514

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.3.5...v0.3.6
Source code(tar.gz)
Source code(zip)
v0.3.5(Aug 2, 2022)
v.0.3.5 - 2022-08-02

What's Changed

🐞 Fix inference for draem by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/470

🐞 🛠 Bug fix in the inferencer by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/475

Source code(tar.gz)
Source code(zip)
v0.3.4(Aug 1, 2022)
What's Changed

Add encoding to LONG_DESCRIPTION in setup.py by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/419

Fix visualization by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/417

Fix openvino circular import issue by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/416

Fix inferener arg names and weight path issue. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/422

Remove the redundant loss_val by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/425

📃 Add documentation for gradio inference by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/427

Add pre_train as a configurable parameter by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/431

🛠 Fix config files and refactor dfkde by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/435

Add metric visualizations by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/429

Fix: data split issue by @jeongHwarr in https://github.com/openvinotoolkit/anomalib/pull/404

🚚 Move perlin noise to common folder by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/424

Support null seed by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/437

🐞 Change if statement by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/439

Fix visualizer for classification, mode=simple by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/442

Feature/aupro test by @ORippler in https://github.com/openvinotoolkit/anomalib/pull/444

Replace PyTorchLightning extras dependency by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/455

🛠 Fix tox configuration by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/446

Ignore ipynb files to detect the repo language by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/456

Move configuration from tox to pyproject by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/458

Add Torch Inferencer and Update Openvino and Gradio Inferencers. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/453

Address markdownlint issues by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/460

🐞 Fix HPO by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/462

Remove docs requirements by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/467

Add codacy badge to readme by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/468

New Contributors

@ORippler made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/417

@jeongHwarr made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/404

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/0.3.3...0.3.4
Source code(tar.gz)
Source code(zip)
0.3.3(Jul 5, 2022)
What's Changed

Move initialization log message to base class by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/363

🚚 Move logging from train.py to the getter functions by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/365

🚜 Refactor loss computation by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/364

📝 Add a technical blog post to explain how to run anomalib. by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/359

📚 Add datamodule jupyter notebooks. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/357

📝 Add benchmarking notebook by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/353

➕ Add PyPI downloads badge to the readme. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/370

Update README.md by @innat in https://github.com/openvinotoolkit/anomalib/pull/382

Create Anomalib CLI by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/378

🛠 Fix configs to remove logging heatmaps from classification models. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/387

Add FastFlow model training testing inference via Anomalib API by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/386

PaDim occasionally NaNs in anomaly map by @VdLMV in https://github.com/openvinotoolkit/anomalib/pull/392

Inference + Visualization by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/390

New Contributors

@innat made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/382

@VdLMV made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/392

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v.0.3.2...0.3.3
Source code(tar.gz)
Source code(zip)
0.2.4a(Jun 14, 2022)

This is a legacy release. Please use the latest release.
Source code(tar.gz)
Source code(zip)
v.0.3.2(Jun 9, 2022)
What's Changed

Refactor AnomalyModule and LightningModules to explicitly define class arguments. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/315

🐞 Fix inferencer in Gradio by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/332

fix too many open images warning by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/334

Upgrade wandb version by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/340

Minor fix: Update folder dataset + notebooks link by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/338

Upgrade TorchMetrics version by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/342

🚀 Set pylint version in tox.ini by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/345

Add metrics configuration callback to benchmarking by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/346

➕ Add FastFlow Model by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/336

✨ Add toy dataset to the repository by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/350

Add DRAEM Model by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/344

📃Update documentation by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/280

🏷️ Refactor Datamodule names by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/354

✨ Add Reverse Distillation by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/343

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v.0.3.1...v.0.3.2
Source code(tar.gz)
Source code(zip)
v.0.3.1(May 17, 2022)
What's Changed

🔧 Properly assign values to dataframe in folder dataset. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/272

➕ Add warnings ⚠️ for inproper task setting in config files. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/274

Updated CHANGELOG.md by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/276

➕ Add long description to setup.py to make README.md PyPI friendly. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/279

✨ Add hash check to data download by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/284

➕ Add Gradio by @julien-blanchon in https://github.com/openvinotoolkit/anomalib/pull/283

🔨 Fix nncf key issue in nightly job by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/238

Visualizer improvements pt1 by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/293

🧪 Fix nightly by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/299

🧪 Add tests for benchmarking script by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/297

➕ add input_info to nncf config when not defined by user by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/307

🐞 Increase tolerance + nightly path fix by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/318

➕ Add jupyter notebooks directory and first tutorial for getting-started by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/292

New Contributors

@julien-blanchon made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/283

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.3.0...v.0.3.1
Source code(tar.gz)
Source code(zip)
v0.3.0(Apr 25, 2022)
What's Changed

🛠 Fix get_version in setup.py to avoid hard-coding version. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/229

🐞 Fix image loggers by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/233

Configurable metrics by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/230

Make OpenVINO throughput optional in benchmarking by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/239

Fix configs to properly use pytorch-lightning==1.6 with GPU by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/234

🔨 Minor fix: Ensure docs build runs only on isea-server by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/245

🏷 Rename --model_config_path to config by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/246

Revert "🏷 Rename --model_config_path to config" by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/247

➕ Add --model_config_path deprecation warning to inference.py by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/248

Add console logger by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/241

Add segmentation mask to inference output by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/242

🛠 Fix broken mvtec link, and split url to fit to 120 by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/264

🛠 Fix mask filenames in folder dataset by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/249

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.2.6...v0.3.0
Source code(tar.gz)
Source code(zip)
v0.2.6(Apr 12, 2022)
What's Changed

✏️ Add torchtext==0.9.1 to support Kaggle environments. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/165

🛠 Fix KeyError:'label' in classification folder dataset by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/175

📝 Added MVTec license to the repo by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/177

load best model from checkpoint by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/195

Replace SaveToCSVCallback with PL CSVLogger by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/198

WIP Refactor test by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/197

🔧 Dockerfile enhancements by @LukasBommes in https://github.com/openvinotoolkit/anomalib/pull/172

🛠 Fix visualization issue for fully defected images by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/194

✨ Add hpo search using wandb by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/82

Separate train and validation transformations by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/168

🛠 Fix docs workflow by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/200

🔄 CFlow: Switch soft permutation to false by default to speed up training. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/201

Return only image, path and label for classification tasks in Mvtec and Btech datasets. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/196

🗑 Remove freia as dependency and include it in anomalib/models/components by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/174

Visualizer show classification and segmentation by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/178

↗️ Bump up pytorch-lightning version to 1.6.0 or higher by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/193

🛠 Refactor DFKDE model by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/207

🛠 Minor fixes: Update callbacks to AnomalyModule by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/208

🛠 Minor update: Update pre-commit docs by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/206

✨ Directory streaming by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/210

✏️ Updated documentation for development on Docker by @LukasBommes in https://github.com/openvinotoolkit/anomalib/pull/217

🏷 Fix Mac M1 dependency conflicts by @dreaquil in https://github.com/openvinotoolkit/anomalib/pull/158

🐞 Set tiling off in pathcore to correctly reproduce the stats. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/222

🐞fix support for non-square images by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/204

Allow specifying feature layer and pool factor in DFM by @nahuja-intel in https://github.com/openvinotoolkit/anomalib/pull/215

📝 Add GANomaly metrics to readme by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/224

↗️ Bump the version to 0.2.6 by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/223

📝 🛠 Fix inconsistent benchmarking throughput/time by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/221

assign test split for folder dataset by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/220

🛠 Refactor model implementations by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/225

New Contributors

@LukasBommes made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/172

@dreaquil made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/158

@nahuja-intel made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/215

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v.0.2.5...v0.2.6
Source code(tar.gz)
Source code(zip)
v.0.2.5(Mar 25, 2022)
What's Changed

Bugfix: fix random val/test split issue by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/48

Fix Readmes by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/46

Updated changelog by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/49

add distinction between image and pixel threshold in postprocessor by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/50

Fix docstrings by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/22

Fix networkx requirement by @LeonidBeynenson in https://github.com/openvinotoolkit/anomalib/pull/52

Add min-max normalization by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/53

Change hardcoded dataset path to environ variable by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/51

Added cflow algorithm by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/47

perform metric computation on cpu by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/64

Fix Inferencer by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/60

Updated readme for cflow and change default config to reflect results by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/68

Fixed issue with model loading by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/69

Docs/sa/fix readme by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/71

Updated coreset subsampling method to improve accuracy by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/73

Revert "Updated coreset subsampling method to improve accuracy" by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/79

Replace SupportIndex with int by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/76

Added reference to official CFLOW repo by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/81

Fixed issue with k_greedy method by @blakshma in https://github.com/openvinotoolkit/anomalib/pull/80

Fix Mix Data type issue on inferencer by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/77

Create CODE_OF_CONDUCT.md by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/86

✨ Add GANomaly by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/70

Reorder auc only when needed by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/91

Bump up the pytorch lightning to master branch due to vulnurability issues by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/55

🚀 CI: Nightly Build by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/66

Refactor by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/87

Benchmarking Script by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/17

🐞 Fix tensor detach and gpu count issues in benchmarking script by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/100

Return predicted masks in predict step by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/103

Add Citation to the Readme by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/106

Nightly build by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/104

c_idx cast to LongTensor in random sparse projection by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/113

Update Nightly by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/126

Updated logos by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/131

Add third-party-programs.txt file and update license by @LeonidBeynenson in https://github.com/openvinotoolkit/anomalib/pull/132

🔨 Increase inference + openvino support by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/122

Fix/da/image size bug by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/135

Fix/da/image size bug by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/140

optimize compute_anomaly_score by using torch native funcrtions by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/141

Fix IndexError in adaptive threshold computation by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/146

Feature/data/btad by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/120

Update for nncf_task by @AlexanderDokuchaev in https://github.com/openvinotoolkit/anomalib/pull/145

fix non-adaptive thresholding bug by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/152

Calculate feature map shape patchcore by @alexriedel1 in https://github.com/openvinotoolkit/anomalib/pull/148

Add transform_config to the main config.yaml file. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/156

Add Custom Dataset Training Support by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/154

Added extension as an option when saving the result images. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/162

Update anomalib version and requirements by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/163

New Contributors

@LeonidBeynenson made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/52

@blakshma made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/47

@alexriedel1 made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/113

@AlexanderDokuchaev made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/145

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v.0.2.4...v.0.2.5
Source code(tar.gz)
Source code(zip)
v.0.2.4(Dec 22, 2021)
What's Changed

Bump up the version to 0.2.4 by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/45

fix heatmap color scheme by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/44

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v.0.2.3...v.0.2.4
Source code(tar.gz)
Source code(zip)
v.0.2.3(Dec 23, 2021)
What's Changed

Address docs build failing issue by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/39

Fix docs pipeline 📄 by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/41

Feature/dick/anomaly score normalization by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/35

Shuffle train dataloader by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/42

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.2.2...v.0.2.3
Source code(tar.gz)
Source code(zip)
v0.2.2(Dec 20, 2021)
What's Changed

Add PR and Issue Templates by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/30

Organize anomalib dependencies by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/32

Limit parallel runners by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/38

Bump lxml from 4.6.3 to 4.6.5 in /requirements by @dependabot in https://github.com/openvinotoolkit/anomalib/pull/37

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.2.1...v0.2.2
Source code(tar.gz)
Source code(zip)
v0.2.1(Dec 16, 2021)
What's Changed

Bump up anomalib version

Docs/dick/root readme by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/31

Add wandb logger by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/23

Full Changelog: https://github.com/openvinotoolkit/anomalib/compare/v0.2.0...v0.2.1
Source code(tar.gz)
Source code(zip)
v0.2.0(Dec 14, 2021)
What's Changed

Address compatibility issues with OTE, that are caused by the legacy code. by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/24

Initial docs string by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/9

Load model did not work correctly as DFMModel did not inherit by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/5

Refactor/samet/data by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/8

Delete make.bat by @samet-akcay in https://github.com/openvinotoolkit/anomalib/pull/11

TorchMetrics by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/7

ONNX node naming by @djdameln in https://github.com/openvinotoolkit/anomalib/pull/13

Add FPS counter to TimerCallback by @ashwinvaidya17 in https://github.com/openvinotoolkit/anomalib/pull/12

New Contributors

@samet-akcay made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/8

@djdameln made their first contribution in https://github.com/openvinotoolkit/anomalib/pull/7

Full Changelog: https://github.com/openvinotoolkit/anomalib/commits/v0.2.0
Source code(tar.gz)
Source code(zip)