Darknet yolo v3 tiny model for Nuts detection. 1 class detection Two models build: small input for h x w of 416 x 416 big input for h x w of 1024 x 768
There needs to be couple of changes:
-
run it with -thresh 0.35
-
Code change in src/image.c :
a) void draw_box_width(image a, int x1, int y1, int x2, int y2, int w, float r, float g, float b)
// instead of w, use 2 or 3. for(i = 0; i < 2; ++i){ draw_box(a, x1+i, y1+i, x2-i, y2-i, r, g, b);
b) void draw_detections(image im, detection *dets, int num, float thresh, char **names, image **alphabet, int classes) // Comment draw label to not display the label name on the output stream // draw_label(im, top + width, left, label, rgb);