Hi,
The input images below are all obtained from the InterHand2.6M test data, because of these output results I am very confused.
Is there anything I need to do for the input image before doing demo? (ex: resize or what?) or there are some restrictions on input images? (ex: hand must be in the middle of the image?) or it just happens that the ones I picked are all fail cases?
Thanks.
intput image :
doing demo by using Single-image reconstruction and get:
And I noticed that the images provided in the demo folder are all square, so I used cv.resize to make the input to be square.
Then this is the result I get :
Another image is also:
Hi, The input images below are all obtained from the InterHand2.6M test data, because of these output results I am very confused. Is there anything I need to do for the input image before doing demo? (ex: resize or what?) or there are some restrictions on input images? (ex: hand must be in the middle of the image?) or it just happens that the ones I picked are all fail cases?
Thanks.
intput image : doing demo by using Single-image reconstruction and get: And I noticed that the images provided in the demo folder are all square, so I used cv.resize to make the input to be square. Then this is the result I get : Another image is also: