A clean, modular implementation of the Yolov7 model family, which uses the official pretrained weights, with utilities for training the model on custom (non-COCO) tasks.
Given 2 images with inverted aspect ratios (image 1: 1000x500, image 2: 500x1000), and a desired image size of 1000x500, the transform A.LongestMaxSize will not pad either image, resulting in a non-batchable image batch. To resolve this, the images should be resized to the largest size that fits within the given dimensions.
Given 2 images with inverted aspect ratios (image 1: 1000x500, image 2: 500x1000), and a desired image size of 1000x500, the transform
A.LongestMaxSize
will not pad either image, resulting in a non-batchable image batch. To resolve this, the images should be resized to the largest size that fits within the given dimensions.