Obtaining Hard Example Configurations

This API is used to read the hard example configuration file. The path of the hard example configuration file is the data directory of the corresponding skill, for example, /home/hilens/skills/***skill_path/data/hardsample_config.json.

• API calling

  hilens.get_hard_sample_config()

• Return value

  Hard example configuration file in JSON format. Example:

  {
  "hard_sample_setting" :
      [
        {
          "camera_names" :["123"],
          "data_count" : 100,
          "datacur_count" : 100,
          "dataset_name" : "dataset-a3ae",
          "dataset_path" : "https://a.b.csss.obs.cn-north-7.ulanqab.huawei.com/nali/",
          "model_algorithm" : "image_classification",
          "model_id" : "073c4c8674164307ae300b713a4a050c",
          "model_name" : "model-framework5",
          "setting_config" :
              {
                  "thr" : 0.5
              }
        }
      ]
  }

  Table 1 Parameters

  Parameter	Description
  camera_names	Camera name. Data of different cameras can be distinguished during data upload.
  data_count	Total number of uploaded images
  datacur_count	Number of images that have been uploaded
  dataset_name	Name of the target dataset
  dataset_path	Target path, which corresponds to the URL of the dataset
  model_algorithm	Algorithm of a model. Generally, the algorithm is classification or detection.
  model_id	Model ID
  model_name	Model name. Data of different models can be distinguished during data upload.
  setting_config	Other settings. Some configurations are custom, such as the threshold.
  thr	Threshold. You can deliver thresholds on the hard example upload configuration page so that different thresholds can be used in different scenarios.