Transfer Learning for Olfactory Object Detection

TL;DR

This paper explores how style and category similarities in datasets affect transfer learning for olfactory object detection, finding that category matching is more crucial than style similarity, and that additional pretraining improves detection performance.

Contribution

It demonstrates that an extra stage of object detection pretraining enhances performance and suggests category similarity is more important than style similarity in transfer learning.

Findings

Additional pretraining stage significantly improves detection accuracy.

Category similarity between datasets is more impactful than style similarity.

Further experiments are needed to confirm the role of style similarity.

Abstract

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.