Glasses detection, classification and segmentation
MIT License
First stable[^1] release for Glasses Detector. Three tasks are supported for processing images with glasses: classification, detection, and segmentation. This includes glasses types and parts, as well as their shadows and eye-area. For all the available features and examples, please check documentation and notebook.
[!NOTE]
Pre-trained weights for sizelarge
are not available as it seems mostmedium
models perform better (less overfit due to tasks being binary). More experiments need to be conducted beforelarge
weights can be released inv1.1.0
.
As before, although the code for training (e.g., datasets, metrics) is part of PIP package, the actual data processing, training, and model analysis scripts are only available when fully cloning the repository. Please check Data and Running sections for details on how to train/test your own models.
[^1]: Kind of stable since large
weights are missing (technically not but they perform worse than medium
)
Published by mantasu over 1 year ago
Fixed version tag (0.1.0
$\to$ v0.1.0
) when the download link of the model weights is created.
Published by mantasu over 1 year ago
This is the initial BETA release of glasses-detector
package. The package structure is complete, however, not all the models are trained. In case custom models are preferred, training and testing scripts are fully ready.
The pre-trained weights are available for these models:
sunglasses-classifier
full-glasses-segmenter
full-sunglasses-segmenter
Note: for classification,
"huge"
category of weights is missing. For full glasses segmentation,"medium"
category of weights is missing. The other 4 categories of weights for both tasks are available.
Please refer to the documentation page for more details.