mmphego/face_mask_detection_openvino

OVMS Configuration

fcabrera23 opened this issue · 1 comments

Hi,

Thanks for sharing this model, it's great. I'm trying to get this model into the OpenVino Model Server - Do you have the model.json configuration parameters (inputs, outputs) for deploying the model in OVMS? For example, this a config file i used for another model

"inputs": [
          {
            "input_name": "data",
            "input_format": "NCHW",
            "color_format": "BGR",
            "target_height": 320,
            "target_width": 544,
            "channels": 3
          }
        ],
      "outputs": [
          {
              "output_name": "detection_out",
              "confidence_threshold": 0.1,
              "classes": {
                  "background": 0.0,
                  "person": 1.0
              },
              "value_index_mapping": {
                  "image_id": 0,
                  "value": 1,
                  "confidence": 2,
                  "x_min": 3,
                  "y_min": 4,
                  "x_max": 5,
                  "y_max": 6
              }
          }
      ],


Thanks,
Francisco