shreelock / gbvs Goto Github PK

Matplotlib outputs the saliency map correctly

however when I output it using
cv2.imshow("Output using cv2.imshow",saliency_map_gbvs)
its outputs is blown out as shown below:

Is this because of resizing or using incorrect color space? Any help would be appreciated :)

When trying to run an image in BMP format, an error occurs, as follows:
Value error：Input contains NaN，infinity or a value too large for dtype（‘float64’）

Hi,
The code works perfectly for most of the images in my dataset. However, it yields this error with few samples, I doubled checked whether there's some problem with the input images, and apparently, this error is only yielded when computing the gbvs.compute_saliency while ittikochneibur.compute_saliency works just fine.

Dear Author：
I want to calculate GBVS map in my thesis. Since I don't have a Matlab license, is it possible to achieve the same functionality by using this code you provided.

Hi. Does your package support the computation of a GBVS map overlayed image as shown below:

Thanks.

Hi there,
I am trying to find the visual saliency of a set of images using the GBVS model (global saliency using np.average(), minimum local saliency using np.min(), and maximum local saliency using np.max(). These 3 numbers are calculated for the saliency_map_gbvs array of each image). Doing so, I realized that the range of saliency instead of being between 0 and 255, varies widely from image to image. For example, the minimum saliency can be -0.5 or -0.9 and the maximum can go up to 294 or 275 depending on the image. Why is this happening? This makes a comparison of saliency across images impossible. Is there a way to normalize the values using a 0-1 range (0 for the lowest saliency and 1 for the highest)?

Hi all,
I tried the example but got this exception.

Is there anything wrong with the build-up?