The authors achieve 84.8% mIoU on the GOOSE test set by combining MaskDINO, RoPE-ViT backbone, CSEC color correction, and quantile-based training data filtering.
Color shift estimation-and- correction for image enhancement,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Technical Report for ICRA 2025 GOOSE 2D Semantic Segmentation Challenge: Leveraging Color Shift Correction, RoPE-Swin Backbone, and Quantile-based Label Denoising Strategy for Robust Outdoor Scene Understanding
The authors achieve 84.8% mIoU on the GOOSE test set by combining MaskDINO, RoPE-ViT backbone, CSEC color correction, and quantile-based training data filtering.