search

oarriaga / paz

Hierarchical perception library in Python for pose estimation, object detection, instance segmentation, keypoint estimation, face recognition, etc.
https://oarriaga.github.io/paz/
MIT License
617 stars 102 forks source link

Efficientdet #230

Closed Manojkumarmuru closed 1 year ago

Manojkumarmuru commented 1 year ago
  1. Dependence of test cases (especially those that check the Efficientdet architecture) on .hdf5/JSON/.npy or any other types of files removed.
  2. Test case to check the functionality of the anchor box building method added. The test includes the following:
    • Checks the range of coordinates of the center of anchor boxes. [Checks if the boxes are valid.]
    • Checks the sign of anchor box width and height. [Checks if the boxes are valid.]
    • Checks if the anchor boxes are symmetrically distributed along the X and Y direction.
    • Checks if the method builds anchor boxes of the given aspect ratio.
    • Checks if a correct number of anchor boxes are built for a given input feature size.