Took subsections of the given image by cropping it into five small parts. Performed data augmentation over those 5 cropped images and then annotated each one them on makesense.ai
Used the YOLOv5 model to train for a total of 100 epochs on the dataset created after augmentation.
It was able to detect a total of 101 transistors in the image, where the true count of the transistors is 100.