Day 27
I tested some more models today- no significant improvements. But I got to meet with both of my advisors in the morning for at least an hour to go over my presentation so I got a lot of good feedback and also some explanations for why the results haven’t been that great.
It turns out that the U-Net structure is ok for RGB-to-thermal translation, which is reflected in the graphs and output images, but not so great for thermal to RGB because that translation ends up with increased complexity/texture. Translating from thermal to RGB accurately is difficult on its own, but with a model and/or loss function not suited to that translation specifically, it could mislead the training. That could also explain why the results using the paired dataset are worse- it’s not making the right connections between the two modalities and when it only has the exact image to compare to rather than something more varied it actually makes it worse. I also got suggestions to improve the image quality like adding a smoothing term for the holes in many of the images (if I didn’t have trypophobia before I certainly do now) and training for longer. It would have been nice to have known all this earlier as I don’t really have enough time now to implement these changes, but I’m going to add it to the “reflection/future works” section of my presentation.
It was also Hannah’s birthday today so we had a celebration at noon. Unfortunately she had already gone out to lunch but we still had a fun time.
I worked more on my presentation and I like it a lot better now. The run-through is tomorrow though and I still need to figure out what to say!
It turns out that the U-Net structure is ok for RGB-to-thermal translation, which is reflected in the graphs and output images, but not so great for thermal to RGB because that translation ends up with increased complexity/texture. Translating from thermal to RGB accurately is difficult on its own, but with a model and/or loss function not suited to that translation specifically, it could mislead the training. That could also explain why the results using the paired dataset are worse- it’s not making the right connections between the two modalities and when it only has the exact image to compare to rather than something more varied it actually makes it worse. I also got suggestions to improve the image quality like adding a smoothing term for the holes in many of the images (if I didn’t have trypophobia before I certainly do now) and training for longer. It would have been nice to have known all this earlier as I don’t really have enough time now to implement these changes, but I’m going to add it to the “reflection/future works” section of my presentation.
It was also Hannah’s birthday today so we had a celebration at noon. Unfortunately she had already gone out to lunch but we still had a fun time.
I worked more on my presentation and I like it a lot better now. The run-through is tomorrow though and I still need to figure out what to say!
Comments
Post a Comment