prediction_ex: prediction_ex.rda

In jthinke/vvipr: Verify VIAME Predictions

prediction_ex.rda

Description

An example prediction data set formatted as a data.frame for use the 'assess_overlap_shiny' function. These data represent the model predictions for a single image and a single class (penguin). The data.frame faithfully represents the original VIAME output file and has headers defined follows:

Usage

prediction_ex

Format

A data frame with 11 columns and 75 rows

Details

• DETECTION_ID. Unique identifier for each annotation

• PIC_NAME. The file name of the image used for annotation

• IMAGE. A numeric index identifying the image

• TLX. Top left X coordinate of bounding box

• TLY. Top left Y coordinate of bounding box

• BRX. Bottom right X coordinate of bounding box

• BRY. Bottom right Y coordinate of bounding box

• CONF. Confidence level of truth annotation. This is 1 by default, since it represents a true observation.

• TARGET. ????

• CLASS. The class identifier for the object to be detected.

• CONF_2. A repeat of the confidence level of the annotation.

jthinke/vvipr documentation built on Nov. 3, 2023, 8:39 p.m.