English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The mature soybean plants are of complex architecture with pods frequentlytouching each other, posing a challenge for in-situ segmentation of on-branchsoybean pods. Deep learning-based methods can achieve accurate training andstrong generalization capabilities, but it demands massive labeled data, whichis often a limitation, especially for agricultural applications. As lacking thelabeled data to train an in-situ segmentation model for on-branch soybean pods,we propose a transfer learning from synthetic in-vitro soybean pods. First, wepresent a novel automated image generation method to rapidly generate asynthetic in-vitro soybean pods dataset with plenty of annotated samples. Thein-vitro soybean pods samples are overlapped to simulate the frequentlyphysically touching of on-branch soybean pods. Then, we design a two-steptransfer learning. In the first step, we finetune an instance segmentationnetwork pretrained by a source domain (MS COCO dataset) with a synthetic targetdomain (in-vitro soybean pods dataset). In the second step, transferring fromsimulation to reality is performed by finetuning on a few real-world maturesoybean plant samples. The experimental results show the effectiveness of theproposed two-step transfer learning method, such that AP$_{50}$ was 0.80 forthe real-world mature soybean plant test dataset, which is higher than that ofdirect adaptation and its AP$_{50}$ was 0.77. Furthermore, the visualizationsof in-situ segmentation results of on-branch soybean pods show that our methodperforms better than other methods, especially when soybean pods overlapdensely.

Transfer Learning from Synthetic In-vitro Soybean Pods Dataset for In-situ Segmentation of On-branch Soybean Pod