I would like to start the topic of precise instance counting in 2D images. It would seem that it is somewhat covered by simple classification using CNNs and/or object detection, however in reality the results provided by both of the former rarely provide an exact enough answer. Are there smart ways to overcome the limitations of the default go-to methods and arrive at accuracy above 99.9% ? This is the question I will try to answer. At first I will just keep adding literature here, before I start writing up some conclusions.
Task Architecture Best Metric References wheat heads CNN MAE 3.85; RMSE 5.19  corn plants YoloV3 Accuracy 0.9866  pistachios RetinaNet + algorithm Accuracy 0.9475  fish CNN Accuracy 0.9755  multiple (a review) multiple (a review) nMAE 0.05 - 0.1  crops, cells, colonies YOLOv3 F1 0.947  plants, UAVs MixNet R² 0.9396 / 0.9875  plant leaves Recurrent Attention SBD 0.849  todo
Automated Counting of Colony Forming Units Using Deep Transfer Learning From a Model for Congested Scenes Analysis - 
Principal Component-Based Image Segmentation -