Export

You are viewing an old version of the documentation. You can switch to the documentation of the latest version by clicking the top-right corner of the page.

You can export a deep learning model and use it in Mech-Vision and Mech-DLK SDK.

Export Model Parameters

Click Export and open the parameter configuration window.

  • Hardware type

    • CPU: Use CPU for deep learning model inference, which will increase inference time and reduce recognition accuracy compared with GPU.

    • GPU (default): Do model inference without optimizing according to the hardware, and the model inference will not be accelerated.

    • GPU (optimization): Do model inference after optimizing according to the hardware. The optimization only needs to be done once and is expected to take 5–15 minutes. The inference time will be reduced after optimization.

  • GPU ID

    The graphics card information of the device deployed by the user. If multiple GPUs are available on the model deployment device, the model can be deployed on a specified GPU.

  • Float precision

    • FP32: high model accuracy, low inference speed.

    • FP16: low model accuracy, high inference speed.

  • Max num of training objects (only visible in the Instance Segmentation module and Object Detection module)

    It refers to the maximum number of training objects during a round of inference, which is 50 by default.

  • Class activation map (CAM) (only visible in the Classification module)

    The inference will slow down when the model saved with CAMs is used in Mech-Vision.

After parameter setting, click Export, select the save path, and wait for the export to complete.

Use Models

Use Models in Mech-Vision

Instructions

The exported models can be used in the Mech-Vision Step Deep Learning Model Package Inference.

Compatibility

  • It is recommended that the models exported from Mech-DLK 2.4.1 or later versions should be used in conjunction with Mech-Vision 1.7.2.

  • Single models exported from Mech-DLK 2.4.1 or later versions can be used in Mech-Vision in version 1.7.0 or above.

  • Single models exported from Mech-DLK 2.4.1 or later versions can only be used in Mech-Vision in version 1.7.2 or above.

  • The cascaded models exported from Mech-DLK 2.4.1 or later versions cannot be used in Mech-Vision 1.7.1 on a CPU-only device.

  • When the Deep Learning Model Package Inference Step performs inference on a model package exported from Mech-DLK 2.2.0 and early versions in Mech-Vision 1.7.2, defect determination rules configured for this model package are invalid. You need to configure them again in Mech-DLK in version 2.4.1 or above before exporting the model package.

  • For model packages of Object Detection exported from Mech-DLK 2.4.1, when the Max num of inference objects is set to 1, and the hardware type of the deep learning model package management tool is CPU, the inference speed becomes very low. It is recommended that the Max num of inference objects be greater than 1.

Click here to view the details on compatibility.

Instance Segmentation

Mech-Vision version Deep learning environment version Mech-Vision Step Mech-DLK version for model Model and configuration file extensions

1.4.0

1.4.0

Instance Segmentation (please start the deep learning server for the Step)

1.4.0

.pth/.py

1.5.x

2.0.0/2.1.0

Instance Segmentation (please start the deep learning server for the Step)

1.4.0

.pth/.py

2.0.0/2.1.0

Instance Segmentation (please start the deep learning server for the Step)

2.0.0/2.1.0

.dlkmp/.dlkcfg

1.6.0

2.0.0/2.1.0

Instance Segmentation (please start the deep learning server for the Step)

1.4.0

.pth/.py

2.0.0/2.1.0

Instance Segmentation (please start the deep learning server for the Step)

2.0.0/2.1.0

.dlkmp/.dlkcfg

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpack

1.6.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.6.2

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.7.0

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.2

No deep learning environment required

Deep Learning Model Package Inference

2.2.0+

.dlkpackC/.dlkpack

Classification

Mech-Vision version Deep learning environment version Mech-Vision Step Mech-DLK version for model Model and configuration file extensions

1.4.0

1.4.0

Image Classification (please start the deep learning server for the Step)

1.4.0

.pth/.json

1.5.x

2.0.0/2.1.0

Image Classification (please start the deep learning server for the Step)

1.4.0

.pth/.json

No deep learning environment required

Deep Learning Inference

2.0.0/2.1.0

.dlkpack

1.6.0

2.0.0/2.1.0

Image Classification (please start the deep learning server for the Step)

1.4.0

.dlkpack

No deep learning environment required

Deep Learning Inference (Mech-DLK 2.1.0/2.0.0)

2.0.0/2.1.0

.dlkpack

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpack

1.6.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.6.2

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.7.0

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.2

No deep learning environment required

Deep Learning Model Package Inference

2.2.0+

.dlkpackC/.dlkpack

Object Detection

Mech-Vision version Deep learning environment version Mech-Vision Step Mech-DLK version for model Model and configuration file extensions

1.4.0

1.4.0

Object Detection (please start the deep learning server for the Step)

1.4.0

.pth/.py

1.5.x

2.0.0/2.1.0

Object Detection (please start the deep learning server for the Step)

1.4.0

.pth/.py

No deep learning environment required

Deep Learning Inference

2.0.0/2.1.0

.dlkpack

1.6.0

2.0.0/2.1.0

Object Detection (please start the deep learning server for the Step)

1.4.0

.dlkpack

No deep learning environment required

Deep Learning Inference (Mech-DLK 2.1.0/2.0.0)

2.0.0/2.1.0

.dlkpack

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpack

1.6.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.6.2

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpackC/.dlkpack

1.7.0

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.2

No deep learning environment required

Deep Learning Model Package Inference

2.2.0+

.dlkpackC/.dlkpack

Defect Segmentation

Mech-Vision version Deep learning environment version Mech-Vision Step Mech-DLK version for model Model and configuration file extensions

1.4.0

1.4.0

Defect Detection (please start the deep learning server for the Step)

1.4.0

.pth/.py

1.5.x

No deep learning environment required

Deep Learning Inference

2.0.0/2.1.0

.dlkpack

1.6.0

No deep learning environment required

Deep Learning Inference (Mech-DLK 2.1.0/2.0.0)

2.0.0/2.1.0

.dlkpack

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpack

1.6.1

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpack

1.6.2

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpack

1.7.0

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.2

No deep learning environment required

Deep Learning Model Package Inference

2.2.0+

.dlkpackC/.dlkpack

Fast Positioning

Mech-Vision version Deep learning environment version Mech-Vision Step Mech-DLK version for model Model and configuration file extensions

1.6.0

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpack

1.6.1

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpack

1.6.2

No deep learning environment required

Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.1+

.dlkpack

1.7.0

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.1

No deep learning environment required

Deep Learning Model Package CPU Inference/Deep Learning Model Package Inference (Mech-DLK 2.2.0+)

2.2.0+

.dlkpackC/.dlkpack

1.7.2

No deep learning environment required

Deep Learning Model Package Inference

2.2.0+

.dlkpackC/.dlkpack

We Value Your Privacy

We use cookies to provide you with the best possible experience on our website. By continuing to use the site, you acknowledge that you agree to the use of cookies. If you decline, a single cookie will be used to ensure you're not tracked or remembered when you visit this website.

  • Preference
  • I accept
  • I don't accept