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

Semantic segmentation is one of the most attractive research fields incomputer vision. In the VIPriors challenge, only very limited numbers oftraining samples are allowed, leading to that the current state-of-the-art anddeep learning-based semantic segmentation techniques are hard to train well. Toovercome this shortcoming, therefore, we propose edge-preserving guidance toobtain the extra prior information, to avoid the overfitting under small-scaletraining dataset. First, a two-channeled convolutional layer is concatenated tothe last layer of the conventional semantic segmentation network. Then, an edgemap is calculated from the ground truth by Sobel operation and followed byconcatenating a hard-thresholding operation to indicate whether the pixel isthe edge or not. Then, the two-dimensional cross-entropy loss is adopted tocalculate the loss between the predicted edge map and its ground truth, termedas an edge-preserving loss. In this way, the continuity of boundaries betweendifferent instances can be forced by the proposed edge-preserving loss.Experiments demonstrate that the proposed method can achieve excellentperformance under small-scale training set, compared to state-of-the-artsemantic segmentation techniques.

Edge-Preserving Guided Semantic Segmentation for VIPriors Challenge