Last active
December 23, 2021 09:30
-
-
Save bavadim/e4fb2f9cc129f344bd40ec001b934a7b to your computer and use it in GitHub Desktop.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Quality benchmarks | |
## search service | |
Dataset: validation part from https://multimediacommons.wordpress.com/yfcc100m-core-dataset/. | |
Methodology of measure: add all images to the Index, request TOP100 for each signature, check if the result is a picture by signature, | |
estimate % of hits. | |
acc TOP100: 0.5 | |
## race classification | |
Dataset: a custom face dataset, 200 images | |
Methodology: standard classification to 5 classes | |
acc: 0.79 | |
## NSF detection | |
Dataset: a custom balanced dataset, 100 images | |
Methodology: standard classification to 5 classes | |
presision: 0.94, recall: 0.9 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment