Kaggle SIIM COVID19 DETECTION

competions-banner This is my repository for attending kaggle SIIM-FISABIO-RSNA COVID-19 Detection

Evaluation

The challenge uses the standard PASCAL VOC 2010 mean Average Precision (mAP) at IoU > 0.5. Note that the linked document describes VOC 2012, which differs in some minor ways (e.g. there is no concept of "difficult" classes in VOC 2010). The P/R curve and AP calculations remain the same. In this competition, we are making predictions at both a study (multi-image) and image level.

Timeline

  • May 17, 2021 - Start Date.
  • August 2, 2021 - Entry Deadline. You must accept the competition rules before this date in order to compete.
  • August 2, 2021 - Team Merger Deadline. This is the last day participants may join or merge teams.
  • August 9, 2021 - Final Submission Deadline (11:59 PM UTC).

My Result (Public & Private Score)

Date Public Private
30/06 0.371 0.341
03/07 0.489 0.484
03/07 0.492 0.484
07/07 0.502 0.504
08/07 0.518 0.506
11/07 0.529 0.524
11/07 0.534 0.528
13/07 0.537 0.523
13/07 0.540 0.532
15/07 0.549 0.547
30/07 0.558 0.547
01/08 0.567 0.557
03/08 0.583 0.574

Place 829th out of 1,324 teams with 91 subs

Dependencies

coming

Task

  • write run_inference.py
    • made some progresses on preprocessing
    • It might take some time to finish including test run
      • notebook was tightly tied to the platform
  • dependencies .yaml
  • files explanation