openvino_inference.py error

[Ishihara-Masabumi]

When I ran openvino_inference.py as instructed in https://github.com/openvinotoolkit/anomalib, the following error occurred.

$ python tools/inference/openvino_inference.py \
>     --config anomalib/models/padim/config.yaml \
>     --weights results/padim/mvtec/bottle/openvino/openvino_model.bin \
>     --meta_data results/padim/mvtec/bottle/openvino/meta_data.json \
>     --input datasets/MVTec/bottle/test/broken_large/000.png \
>     --output results/padim/mvtec/bottle/images
/home/ishihara/anomalib/anomalib/config/config.py:274: 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(
Traceback (most recent call last):
  File "tools/inference/openvino_inference.py", line 108, in <module>
    infer()
  File "tools/inference/openvino_inference.py", line 82, in infer
    inferencer = OpenVINOInferencer(
  File "/home/ishihara/anomalib/anomalib/deploy/inferencers/openvino_inferencer.py", line 52, in __init__
    self.input_blob, self.output_blob, self.network = self.load_model(path)
  File "/home/ishihara/anomalib/anomalib/deploy/inferencers/openvino_inferencer.py", line 78, in load_model
    network = ie_core.read_network(xml_path, bin_path)
  File "ie_api.pyx", line 367, in openvino.inference_engine.ie_api.IECore.read_network
  File "ie_api.pyx", line 401, in openvino.inference_engine.ie_api.IECore.read_network
Exception: Path to the model results/padim/mvtec/bottle/openvino/openvino_model.xml doesn't exist or it's a directory

[samet-akcay]

Can you share your config file. Similar to #854, you may not have trained the model yet, which should generate the openvino model. (Note that to generate the openvino model, you need to configure the config file as instructed on the README.md file.

[Ishihara-Masabumi]

My config file is as follows:

dataset:
  name: mvtec
  format: mvtec
  path: ./datasets/MVTec
  category: bottle
  task: segmentation
  train_batch_size: 32
  eval_batch_size: 32
  num_workers: 8
  image_size: 256 # dimensions to which images are resized (mandatory)
  center_crop: null # dimensions to which images are center-cropped after resizing (optional)
  normalization: imagenet # data distribution to which the images will be normalized: [none, imagenet]
  transform_config:
    train: null
    eval: null
  test_split_mode: from_dir # options: [from_dir, synthetic]
  test_split_ratio: 0.2 # fraction of train images held out testing (usage depends on test_split_mode)
  val_split_mode: same_as_test # options: [same_as_test, from_test, synthetic]
  val_split_ratio: 0.5 # fraction of train/test images held out for validation (usage depends on val_split_mode)
  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
  pre_trained: true
  layers:
    - layer1
    - layer2
    - layer3
  normalization_method: min_max # options: [none, min_max, cdf]

metrics:
  image:
    - F1Score
    - AUROC
  pixel:
    - F1Score
    - AUROC
  threshold:
    method: adaptive #options: [adaptive, manual]
    manual_image: null
    manual_pixel: 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: 42
  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
  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
  sync_batchnorm: false
  tpu_cores: null
  track_grad_norm: -1
  val_check_interval: 1.0 # Don't validate before extracting features.

[Ishihara-Masabumi]

I understood that python means v2.7 and python3 means v3.8 on my PC.
Thanks.

[samet-akcay]

I understood that python means v2.7 and python3 means v3.8 on my PC. Thanks.

To avoid such issues, I highly recommend you use virtual environments.

Regarding your config file, you need to change the following section

optimization:
  export_mode: null #options: onnx, openvino

to

optimization:
  export_mode: openvino #options: onnx, openvino

If export_mode is not openvino, anomalib cannot create an openvino model.

[samet-akcay]

Since this is not an issue, I'm moving this to Q&A in discussions. Feel free to ask more questions from there.