Multi-Scale and sliding window in caffe for object detection and localization.

[Shashi Kant]

Hi, I am using caffe for object detection and localisation. I am cutting the image into a number of images as sliding window at different scale and giving these list of image to caffe for classification. But this method is a time consuming and its not an efficient method. Is there any method through which I can give whole image as input to caffe and it detect the object using sliding window at different scale? Please help if any one has any idea how to perform this... thanks in advance for helping...

[Nickolay Grin]

Hi, you may try to use https://github.com/rbgirshick/py-faster-rcnn or https://github.com/zhaoweicai/mscnn foth such task.
вторник, 10 января 2017 г., 20:26:10 UTC+2 пользователь Shashi Kant написал:

[Stefano Imoscopi]

On Tuesday, 10 January 2017 19:26:10 UTC+1, Shashi Kant wrote:

Hi, I am using caffe for object detection and localisation. I am cutting the image into a number of images as sliding window at different scale and giving these list of image to caffe for classification. But this method is a time consuming and its not an efficient method. Is there any method through which I can give whole image as input to caffe and it detect the object using sliding window at different scale? Please help if any one has any idea how to perform this... thanks in advance for helping...

Also the SSD detector from Google. It is more recent than RCNN. 

https://github.com/weiliu89/caffe/tree/ssd