Unified Insect Image Dataset

Read full text
Authors: DENG DENG
Cited by: 0

Abstract

We build the first dataset for unified insect image segmentation that includes both salient and camouflaged categories captured in diverse real-world natural scenes. Compared to existing datasets, ours features greater taxonomic diversity, higher-resolution images, and fine-grained manual annotations, providing a challenging benchmark for unified segmentation tasks.

Publication details

Journal / source
Mendeley Data
Publication date
2 January 2026
Document type
Dataset