pixellib.whl

This is the release that contains information about different versions of pixellib packages.
0.7.1: Fixed a bug.

0.7.0: PixelLib Pytorch Version with the following new features;

• PointRend is used for segmentation of objects in images and videos.
• Supports extraction of objects from their bounding boxes' coordinates and masks' values.
• Faster and more accurate than the tensorflow version:
  It achieves 0.26 seconds for processing a single image and 4fps for live camera feeds.
  Using A TargetSize of 667 * 447: It achieves 0.20 seconds for processing a single image and 6fps for live camera feeds.
  Using A TargetSize of 333 * 200: It achieves 0.15 seconds for processing a single image and 9fps for live camera feeds.

0.6.6: Added support for the following features;

• Batch image segmentation.

• Ability to change the threshold for performing trained model evaluation.

• Ability to change the size and thickness of the label names and bounding boxes for visualization of segmented images.

0.6.1: Fixed a bug.

0.6.0: It provides support for extraction of segmented objects in video files and live camera feeds.

0.5.5: It provides support for extraction of segmented objects in images and the ability to filter coco model detections to segment a user's target class.

0.5.2: Added the ability to return the polygon points' values of masks.

0.4.9: Added the ability to choose the inference speed mode for instance segmentation.

0.4.8: It provides the ability to change the background of video files and live camera feeds.

0.4.0: It provides the ability to change the background of images.

0.3.0: It provides support for custom training.

0.2.1: Fixed a bug.

0.2.0: It provides support for segmentation of objects in video files, live camera feeds and semantic segmentation of 150 classes of objects.

0.1.0: It provides support for semantic segmentation of 20 classes of objects and instance segmentation of 80 classes of objects.