volvo-cars / FB-SSEM-dataset

The FB-SSEM dataset is a synthetic datasetconsisting of surround-view fisheye camera images and BEV maps from simulated sequences of ego car motion
Creative Commons Attribution Share Alike 4.0 International
33 stars 7 forks source link

when will you release the style transfer tools and process #6

Open Light-- opened 5 months ago

Light-- commented 5 months ago

thanks for sharing this project, i also have real world fisheye images to predict but got no GT, may i ask when will you release the style transfer tools and process?

as mentioned in the paper,

We run neural style transfer [41] on the simulated FB-SSEM dataset images using samples of realworld images as style images. Additionally, we note that the simulated fisheye cameras differ from those mounted on the actual car in terms of the valid regions of the image (i.e., regions not covered by the camera or the ego car). We account for this difference during training by predefining the valid region for every camera in the training data to be the same as that of the corresponding real-world camera.

ektas0330 commented 4 months ago

We used the neural style algorithm described in this paper; the code is available here.