First Time Through (FTT) metric (in percentage) shows how many items labeler annotated right the first time (i.e. reviewer accepted his work on first round). This metric helps to find accurate labelers (i.e. reviewers accepts many of their annotations).
FTT (%) = ACCEPTED * 100 / LABELED
Application shows the table of all labeling jobs in the team that were assigned to specific member and calculates FTT metric for every job (last column). Also it produces average FTT for all non-zero values.
Log in to the team, then go to Ecosystem->Apps page. Find app and press Get button. Now app is added to your team.
Enable checkbox if you want to use also archived (disabled) jobs.
Done!
Apps calculates FTT metric for all jobs. If job has rejected images, these images can be separated to another new labeling job by clicking Restart rejected button. And then labeler will correct his mistakes.
The problem here is that supervisely do not track such new jobs and AVG FTT metric will be slightly incorrect. In best case such new jobs (Restart rejected) have to be ignored.
Now Application provides FTT for every job. And it helps to get main idea how labeler performs. To distinguish jobs (created with Restart rejected button) from the original ones we recommend the following workaround: to name labeling jobs with suffix round-XX (for example labeling job #01 round-01, labeling job #01 round-02, etc...). But in the future versions of the app it will be fixed and the metric will be completely correct.
To see history of runs go to Apps page, click to applications sessions. In front of every session you can see buttons (View and Logs). Press View button to open stopped application session in Read Only mode.