POST
/
projects
/
{project_id}
/
runs
/
{run_id}
/
split
/
{split}
/
insights
/
semseg
/
metrics

Authorizations

Galileo-API-Key
string
headerrequired

Path Parameters

project_id
string
required
run_id
string
required
split
enum<string>
required
Available options:
training,
validation,
test,
inference

Query Parameters

inference_name
string
default:
scope
string | null
default:

Body

application/json
task
string | null
filter_params
object
compare_to
enum<string> | null
Available options:
training,
validation,
test,
inference
map_threshold
number
default: 0.5
meta_cols
string[] | null

Response

200 - application/json

The main metric for Semantic Segmentation.

IoU, or Intersection over Union, is an Semantic Segmentation metric that provides you with a general sense of the performance of your model.

IoU is defined as the intersection of the predicted and ground truth masks Mean IoU is the average IoU across all classes Boundary IoU is the IoU for the boundary of the mask, which protects against bias towards larger masks https://learnopencv.com/intersection-over-union-iou-in-object-detection-and-segmentation/

mean_iou
number
required
boundary_iou
number
required
dice_coefficient
number
required
mean_iou_per_class
object
required

A class to represent a basic bar chart.

labels: List[str] the x axis labels values: List[int | float] the counts for each bar

boundary_iou_per_class
object
required

A class to represent a basic bar chart.

labels: List[str] the x axis labels values: List[int | float] the counts for each bar

dice_per_class
object
required

A class to represent a basic bar chart.

labels: List[str] the x axis labels values: List[int | float] the counts for each bar