Noam Bressler

## Background In order to provide extra value in performance & drift checks, we want to always display information about these quantities in an aggregate manner (as done in the...

[FEAT] Update custom vision task to note that Segmentation is already part of the package

1

**Is your feature request related to a problem? Please describe.** As segmentation is added as a built in part of the package, changes should be introduced to the existing segmentation...

Create new custom task guide - this time for pose estimation

1

Following the completion of #1039, we'll need a new custom task guide to replace the existing one for segmentation. Suggestion - the new guide will be for a pose estimation...

[FEAT] Prediction drift on probabilities for vision data

2

Modify the prediction drift in vision to match what was done in tabular here https://github.com/deepchecks/deepchecks/issues/1775

[FEAT] Enable sorting displayed features in drift by combination of FI and drift score

2

## DoD Add an option to the "sort_feature_by" parameter an "weighted" option that displays the features by the vector multiplication of FI and drift score. Note this option in the...

[FEAT] Resolve two ways of passing n_samples to vision suites / checks

1

**Is your feature request related to a problem? Please describe.** Currently there are 2 ways to sample when running vision checks: 1. Through the init, with a parameter named `sample_size.`...

[BUG] AdaBoost in iris_split_dataset_and_model has different model predictions locally and in test environment

1

**Describe the bug** When using the iris_split_dataset_and_model fixture, the model predictions are different between the local environment and the testing environment. This is result of different predictions of the AdaBoost...

[FEAT] Replace VisionSegmentPerformance and ModelErrorAnalysis with Vision Weak Segment Performance

1

**Is your feature request related to a problem? Please describe.** In the tabular package we replaced SegmentPerformance and ModelErrorAnalysis with the new Weak Segment Performance. Should do also for vision....

Further improvements to custom check guide

1

## Background There have been a few suggestions for further improving the structure custom check guide: 1. Have a Customization section also in tabular and vision. The index of the...

**Is your feature request related to a problem? Please describe.** Following https://github.com/deepchecks/deepchecks/issues/1724, it seems that the source of long run times in Trust Score computation is the OHE that creates...