Training metrics

To view your project’s training metrics, go to Neural network and select the training model. Select Download to download the training metrics.

Instance segmentation for Pixel and Vector projects

Metric

Description

Possible range

Total loss

The current value of the loss function as described in Mask RCNN, a paper by Kalming He et. al.

Any range

Mask loss

The current value of the component of the total loss function that is responsible for the mask head.

Any range

Classes loss

The current value of the component of the total loss function that is responsible for the classes head.

Any range

Bounding box loss

The current value of the component of the total loss function that is responsible for the bounding box head.

Any range

Bounding box mAP

The mean average precision of the bounding boxes over all the training classes evaluated on the validation set.

0-100

Bounding box mAP at IoU=0.50

The mean average precision of bounding boxes at IoU=0.5 evaluated on the validation set.

0 - 100 (per each IoU)

Bounding box mAP at IoU=0.75

The mean average precision of bounding boxes at IoU=0.75 evaluated on the validation set.

0 - 100 (per each IoU)

Bounding box mAP for <class>

The average precision of bounding boxes found for particular class, i.e., how precise the model was for a given class. It is reported for all the training classes.

0-100

Segmentation mAP

The mean average precision of the segmentation over all the training classes.

0-100

Segmentation mAP at IoU=0.50

The mean average precision of segmentation at IoU=0.5

0-100

Segmentation mAP at IoU=0.75

Mean average precision of segmentation at IoU=0.75

0-100

Segmentation mAP for <class>

The average precision of the segmentation found for a particular class, i.e., how precise the model was for a given class.

0-100

ETA

Estimated time of arrival

Object detection for Vector projects

Metric

Description

Possible range

Total loss

The current value of the loss function as described in Mask RCNN, a paper by Kalming He et. al.

Any range

Classes loss

The current value of the component of the total loss function that is responsible for the classes head.

Any range

Bounding box loss

The current value of the component of the total loss function that is responsible for the bounding box head.

Any range

Bounding box mAP

The mean average precision of the bounding boxes over all the training classes evaluated on the validation set.

0-100

Bounding box mAP at IoU=0.50

The mean average precision of the bounding boxes over all the training classes at IoU=0.50 evaluated on the validation set.

0 - 100 (per each IoU)

Bounding box mAP at IoU=0.75

The mean average precision of the bounding boxes over all the training classes at IoU=0.75 evaluated on the validation set.

0 - 100 (per each IoU)

Bounding box mAP for <class>

The average precision of the bounding boxes found for a particular class, i.e., how precise the model was for a given class. It is reported for all the training classes.

0-100

ETA

Estimated time of arrival

Keypoint detection

Metric

Description

Possible range

Total loss

The current value of the loss function as described in Mask RCNN, a paper by Kalming He et. al.

Any range

Keypoint loss

The current value of the component of the total loss function that is responsible for the key points head.

Any range

Classes loss

The current value of the component of the total loss function that is responsible for the classes head.

Any range

Bounding box loss

The current value of the component of total loss function that is responsible for the bounding box head.

Any range

Key point mAP

The mean average precision of the key points over all the key point types evaluated on the whole validation set.

0-100

mAP at OKS=0.50

The mean average precision of keypoints at OKS=0.5 evaluated on the whole validation set.

0 - 100 (per each IoU)

mAP at OKS=0.75

The mean average precision of bounding boxes at IoU=0.75 evaluated on the whole validation set.

0 - 100 (per each IoU)

Bounding box mAP

The mean average precision of the bounding boxes over all training classes. It is evaluated on the whole validation set.

0-100

Bounding box mAP at IoU=0.50

The mean average precision of bounding boxes at IoU=0.5 evaluated on the whole validation set.

0 - 100 (per each IoU)

Bounding box mAP at IoU=0.75

The mean average precision of bounding boxes at IoU=0.75 evaluated on the whole validation set.

0 - 100 (per each IoU)

ETA

Estimated time of arrival