Pseudo-Label Refinement for Robust Wheat Head Segmentation via Two-Stage Hybrid Training

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Jiang, Jiahao, Yang, Zhangrui, Wang, Xuanhan, Song, Jingkuan
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917143919984640
author Jiang, Jiahao
Yang, Zhangrui
Wang, Xuanhan
Song, Jingkuan
author_facet Jiang, Jiahao
Yang, Zhangrui
Wang, Xuanhan
Song, Jingkuan
contents This extended abstract details our solution for the Global Wheat Full Semantic Segmentation Competition. We developed a systematic self-training framework. This framework combines a two-stage hybrid training strategy with extensive data augmentation. Our core model is SegFormer with a Mix Transformer (MiT-B4) backbone. We employ an iterative teacher-student loop. This loop progressively refines model accuracy. It also maximizes data utilization. Our method achieved competitive performance. This was evident on both the Development and Testing Phase datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2512_11874
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Pseudo-Label Refinement for Robust Wheat Head Segmentation via Two-Stage Hybrid Training
Jiang, Jiahao
Yang, Zhangrui
Wang, Xuanhan
Song, Jingkuan
Computer Vision and Pattern Recognition
This extended abstract details our solution for the Global Wheat Full Semantic Segmentation Competition. We developed a systematic self-training framework. This framework combines a two-stage hybrid training strategy with extensive data augmentation. Our core model is SegFormer with a Mix Transformer (MiT-B4) backbone. We employ an iterative teacher-student loop. This loop progressively refines model accuracy. It also maximizes data utilization. Our method achieved competitive performance. This was evident on both the Development and Testing Phase datasets.
title Pseudo-Label Refinement for Robust Wheat Head Segmentation via Two-Stage Hybrid Training
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.11874