Timeline for How to deal with unbalanced data in pixelwise classification?
Current License: CC BY-SA 4.0
5 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 27, 2018 at 20:24 | vote | accept | akrasuski1 | ||
| May 27, 2018 at 20:24 | answer | added | akrasuski1 | timeline score: 0 | |
| May 23, 2018 at 2:03 | comment | added | Vladislav Gladkikh | I think that your artificial example does not quite correspond to your real task. The roads are continuous lines. However, the zeros in your example are isolated. This makes the problem more difficult. You real task is the task of line detection. There are many methods in image processing for that. Maybe some of them use CNN. I think it is worth changing you artificial example, make it contain continuous lines with intersections, etc just like roads. This will make the task easier which will increase a chance that someone comes up with a solution. | |
| May 17, 2018 at 19:23 | review | First posts | |||
| May 18, 2018 at 1:45 | |||||
| May 17, 2018 at 19:21 | history | asked | akrasuski1 | CC BY-SA 4.0 |