Logo-Detection-YOLOv2
Logo-Detection-YOLOv2 copied to clipboard
akarshzingade
What is weight file size ??
@akarshzingade thanks for giving an idea for the implementation . wht is the size of the file generated ??? can we use this dataset to check the orientation of an logo
Well, the weight file linked in this repository is 275MB.
I am not sure about the orientation part.
@akarshzingade even after using yolo v2 the size of the model is around 275 mb can we minimize the model size ???
After detecting the logo i needed to check the orientation of the logo , ie to say where its horizontal or vertical or at any other angle is it jpossible to identify this
thanking you in advance
Maybe you can use different architectures for YOLOv2- like tinyYOLO. It has a lower memory footprint.
Once you get the bounding box for your logo, you can use orientation detection techniques to find the orientation. A quick glance google showed this: https://github.com/akarshzingade/Logo-Detection-YOLOv2/issues/2#issuecomment-366869826. I hope it gives you a good starter.
help me in the instruction of obj.names i didnot understand this instruction,!! 3.Path to store the train.txt/test.txt file and obj.names file. 4.Name of the text file to store the paths to the images for train/test.
@mathan1312 these are the same steps in the yolov2 training where the test and train files in the txt format has all the list of file which has to be trained
yeah @abhigoku10 , I think i given the path for input folder , then output folder . thereafter it created some of images and .text files. i don't know what to do further for testing my own images
How do I chnage the predicted labels if I am using the pretrained logo detection model?
On running
./darknet detect cfg/yolov2_logo_detection.cfg YOLOv2_logo_detection_10000th_iteration.weights data/logo.jpg
for the image attached I get
data/logo.jpg: Predicted in 7.145362 seconds. wine glass: 77%
which is not the correct label (and not present in label set in obj.names)
