Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Output improvements for detector results: #741

Merged
merged 2 commits into from
May 8, 2018

Conversation

IlyaOvodov
Copy link

When printing detector results, output was done in random order, obfuscating results for interpreting. Now:

  1. Text output includes coordinates of rects in (left,right,top,bottom in pixels) along with label and score
  2. Text output is sorted by rect lefts to simplify finding appropriate rects on image
  3. If several class probs are above threshold for some detection, the most probable one is written first and coordinates for others are not repeated
  4. Rects are imprinted in image in order from low to high probability, so most probable rects are always on top and are not overlayed by less probable ones.
  5. Most probable label for rect is always written first
    Also:
  6. Message about low GPU memory includes required memory size

When printing detector results, output was done in random order, obfuscating results for interpreting. Now:
1. Text output includes coordinates of rects in (left,right,top,bottom in pixels) along with label and score
2. Text output is sorted by rect lefts to simplify finding appropriate rects on image
3. If several class probs are > thresh for some detection, the most probable is written first and coordinates for others are not repeated
4. Rects are imprinted in image in order by their best class prob, so most probable rects are always on top and not overlayed by less probable ones
5. Most probable label for rect is always written first
Also:
6. Message about low GPU memory include required amount
When printing detector results, output was done in random order, obfuscating results for interpreting. Now:
1. Text output includes coordinates of rects in (left,right,top,bottom in pixels) along with label and score
2. Text output is sorted by rect lefts to simplify finding appropriate rects on image
3. If several class probs are > thresh for some detection, the most probable is written first and coordinates for others are not repeated
4. Rects are imprinted in image in order by their best class prob, so most probable rects are always on top and not overlayed by less probable ones
5. Most probable label for rect is always written first
Also:
6. Message about low GPU memory include required amount

Upd:
* struct detection changes reverted to keep it unchanged
* cmd line key "-ext_output" is added to manage extended output (rect coords in detector test mode)

# Conflicts:
#	src/detector.c
@IlyaOvodov
Copy link
Author

Upd:

  • struct detection changes reverted to keep it unchanged
  • cmd line key "-ext_output" is added to manage extended output (rect coords in detector test mode)

@AlexeyAB AlexeyAB merged commit 0948df5 into AlexeyAB:master May 8, 2018
@AlexeyAB
Copy link
Owner

AlexeyAB commented May 8, 2018

@IlyaOvodov Thanks!

@AlexeyAB
Copy link
Owner

AlexeyAB commented May 8, 2018

Some problems: #770

@IlyaOvodov IlyaOvodov deleted the Fix_detector_output branch May 14, 2018 08:38
jktzes pushed a commit to jktzes/darknet that referenced this pull request Jul 22, 2018
Output improvements for detector results:
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants