Timeline for Improving segmentation diffusion
Current License: CC BY-SA 3.0
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| May 15, 2014 at 13:24 | comment | added | Silouane Gerin | Finally got somewhat satisfying results. I thought I'd let you know. Thanks again for your help geometrikal. | |
| May 12, 2014 at 12:37 | comment | added | geometrikal | Thanks! If you solve it please put an answer with the solution as this is a very interesting technique. :) | |
| May 12, 2014 at 12:33 | comment | added | Silouane Gerin | Ok thanks. I'm going to try those ideas. I gave you the bounty even if my problem isn't solved but I really appreciate the time you took to answer me. | |
| May 12, 2014 at 12:32 | history | bounty ended | Silouane Gerin | ||
| May 10, 2014 at 1:00 | comment | added | geometrikal | How about enforce a direction for the graph through the layers. Once a path has gone up or down, label it as an up or down path and don't let it reverse direction. Perhaps that will stop it going up through the layers, finding a leak, and then going back down. Other ideas: add weighting for shared boundary length (single layer) or shared area (across layers). | |
| May 9, 2014 at 7:16 | comment | added | Silouane Gerin | Thanks again for your interest. I got 2D superpixels that are linked in 3D. Unfortunately I don't have an efficient way to debug it as it is really difficult for me to run through the algorithm with such big graphs (300 000 nodes, 2M edges). | |
| May 8, 2014 at 13:34 | comment | added | geometrikal | Sounds difficult, are the super pixels in 3D or does the path join 2D super pixels on each layer? (I haven't really done much in 3D). | |
| May 7, 2014 at 13:38 | comment | added | Silouane Gerin | However if I don't try to process the whole 3D volume but focus only on a slice, things appears fine without weird labeling. So I guess my problems come from trying to apply the algorithm in 3D which may cause leakage. | |
| May 7, 2014 at 12:18 | comment | added | Silouane Gerin | Thanks for your answer. I implemented your idea but it didn't seem to solve the core problems I had: some clicks (labeling input) still produce incoherent labeling. For example if I click somewhere to indicate that one region is background, another superpixel or region, further away and not directly linked to the clicked region/superpixel may be labeled as object. I used a combination of color distance and edge strength (gradient magnitude right?) for the weight of my nodes. | |
| May 5, 2014 at 21:21 | history | answered | geometrikal | CC BY-SA 3.0 |