Filter Detections Module

NVIDIA Docs Hub NVIDIA Morpheus NVIDIA Morpheus (23.07) Filter Detections Module

Filter message by a classification threshold.

The Filter Detections module is used to filter rows from a dataframe based on values in a tensor using a specified criteria. Rows in the meta dataframe are excluded if their associated value in the probs array is less than or equal to threshold.

Configurable Parameters

Parameter	Type	Description	Example Value	Default Value
`copy`	boolean	Whether to copy the rows or slice them	true	`true`
`field_name`	string	Name of the field to filter on	“probs”	`probs`
`filter_source`	string	Source of the filter field	“AUTO”	`AUTO`
`schema`	dictionary	Schema configuration	See Below	`-`
`threshold`	float	Threshold value to filter on	0.5	`0.5`

schema

Key	Type	Description	Example Value	Default Value
`encoding`	string	Encoding	“latin1”	`latin1`
`input_message_type`	string	Pickled message type	“pickle_message_type”	`[Required]`
`schema_str`	string	Schema string	“string”	`[Required]`

Example JSON Configuration

Copy
Copied!

            
            {
  "field_name": "probs",
  "threshold": 0.5,
  "filter_source": "AUTO",
  "copy": true,
  "schema": {
    "input_message_type": "pickle_message_type",
    "encoding": "utf-8"
  }
}

Default Settings

Property	Value
copy	False
field_name	probs
filter_source	AUTO
threshold	0.5