Transfer Learning for Olfactory Object Detection
This work addresses improving object detection accuracy for researchers, but it is incremental as it builds on existing transfer learning methods.
The study examined how style and category similarity in pretraining datasets affect object detection performance, finding that an additional object-detection pretraining stage significantly boosts detection results, with category matching appearing more important than style similarity.
We investigate the effect of style and category similarity in multiple datasets used for object detection pretraining. We find that including an additional stage of object-detection pretraining can increase the detection performance considerably. While our experiments suggest that style similarities between pre-training and target datasets are less important than matching categories, further experiments are needed to verify this hypothesis.