Multi-Task Partially Supervised Learning for Super-Resolution and Semantic Segmentation on Earth Observation data

Hoàng-Ân Lê; Minh-Tan Pham; Solange Lemai-Chenevier; Daniel Greslou

cs.CVOct 5, 2026

Multi-Task Partially Supervised Learning for Super-Resolution and Semantic Segmentation on Earth Observation data

Authors: Hoàng-Ân Lê, Minh-Tan Pham, Solange Lemai-Chenevier, Daniel Greslou

Organizations: Universit´e Bretagne Sud, IRISA, UMR 6074, Vannes, France · Centre National d’Etudes Spatiales (CNES), Toulouse, France

Abstract

Super-resolution and semantic segmentation are known to benefit one another, especially in the Earth observation context. However, learning both tasks in a joint model often requires both task annotations, which is impractical and expensive. In this paper, we study the multi-task partially supervised learning paradigm for both tasks, where each example is assumed to have only a single-task annotation. To that end, we examine two multi-task architectural variations, the sequential and shared variants, and then propose a hybrid variant and a re-projection loss to benefit from the shared representation and enforce image quality of super-resolution when training with semantic segmentation. Experiments show favorable results compared to the SOTA sequential variant. Source code will be published at https://github.com/lhoangan/munera.

Figures & tables

Panel of Figure 1 — Figure 1: The multi-task architecture (color-coded) with a shared encoder (a), its sequential variant (b), and the hybrid variant (c) for 2 tasks, super-resolution and semantic segmentation . The encoder consists of $B=23$ residual-in-residual blocks [ 14 ] , the super-resolution decoder comprises 2 upsampling followed by 2 convolution layers, and the segmentation decoder is taken from HRNet-W18 architecture [ 15 ]

	variant	SSIM	PSNR
SS-guided SR [ 4 ]	sequential	0.746	23.890
SR4IR [ 2 ]	sequential	0.694	22.830
Ours	shared	0.804	25.491
Ours	hybrid	0.804	25.446

Table 1: Comparing our shared and hybrid MTL variants with the state-of-the-art methods employing the sequential variant for super-resolution on the Vaihingen test set. Our methods achieve better fidelity quality.

	variant	Vaihingen	Pleiades
SS-guided SR [ 4 ]	sequential	0.597	0.492
SR4IR [ 2 ]	sequential	0.716	0.520
Ours	shared	0.605	0.508
Ours	hybrid	0.721	0.531

Table 2: Comparing our shared and hybrid MTL variants with the state-of-the-art methods employing the sequential variant for semantic segmentation on both datasets. Our methods perform on par with the others on segmentation.

Panel of Figure 3 — Figure 3: Qualitative results on the Vaihingen test. The shared and hybrid variants show richer in textures compared to the overly-smooth SS-guided SR [ 4 ] and the blurry SR4IR [ 2 ] .

Panel of Figure 4 — Figure 4: Qualitative results on the Pléiades test set. Our methods approach the photometric consistency with the referenced bicubic, but are sharper, while the SOTAs show color leaks and are slightly blurrier. SR4IR [ 2 ] has a strong white-balance bias.

		sequential
with GAN	with $\mathcal{L}_{\text{rpj}}$	SSIM	PSNR
✓	✓	0.796	26.000
		0.764	23.835
	✓	0.821	26.491

Table 3: Ablation study of GAN loss and re-projection loss

\mathcal{L}_{\text{rpj}}

. Removing GAN loss while adding re-projection loss reduces super-resolution distortion.

Explore similar work

Beyond Visual Fidelity: Benchmarking Super-Resolution Models for Large-Scale Remote Sensing Imagery via Downstream Task Integration
May 1, 2026Zhili Li, Kangyang Chai, Zhihao Wang +6Hyperspectral Image Super-Resolution Remote Sensing
University of Maryland · University of Pittsburgh · Worcester Polytechnic Institute +1
MapSR: Prompt-Driven Land Cover Map Super-Resolution via Vision Foundation Models
Apr 16, 2026Ruiqi Wang, Qi Yu, Jie Ma +1Land Cover Dense Annotation
School of Information Science and Technology, Beijing Foreign Studies University, Beijing 100089, China
Simon-SR: Spatially Adaptive Modulation and Visual Prompt Adaptation for Text-Reinforced Super-Resolution
Jul 10, 2026Haotong Cheng, Yuxuan Li, Zijie CuiReal-World Image Super-Resolution Generative Super-Resolution
College of Electronic Science and Engineering, Jilin University, Changchun China

May 1, 2026, cs.CVJ/K move · Enter open · S save

Beyond Visual Fidelity: Benchmarking Super-Resolution Models for Large-Scale Remote Sensing Imagery via Downstream Task Integration

Zhili Li, Kangyang Chai, Zhihao Wang, Xiaowei Jia+5

University of Maryland · University of Pittsburgh · Worcester Polytechnic Institute+1

Hyperspectral Image Super-Resolution Remote Sensing Visual Fidelity

Open paper AskPDF arXiv

Super-resolution (SR) techniques have made major advances in reconstructing high-resolution images from low-resolution inputs. The increased resolution provides visual enhancement and utility for monitoring tasks. In particular, SR has been increasingly developed for satellite-based Earth observation, with applications in urban planning, agriculture, ecology, and disaster response. However, existing SR studies and benchmarks typically use fidelity metrics such as PSNR or SSIM, whereas the true utility of super-resolved images lies in supporting downstream tasks such as land cover classification, biomass estimation, and change detection. To bridge this gap, we introduce GeoSR-Bench, a downstream task-integrated SR benchmark dataset to evaluate SR models beyond fidelity metrics. GeoSR-Bench comprises spatially co-located, temporally aligned, and quality-controlled image pairs from about 36,000 locations across diverse land covers, spanning resolutions from 500m to 0.6m. To the best of our knowledge, GeoSR-Bench is the first SR benchmark that directly connects improved image resolution from SR models with downstream Earth monitoring tasks, including land cover segmentation, infrastructure mapping, and biophysical variable estimation. Using GeoSR-Bench, we benchmark GAN, transformer, neural operator, and diffusion-based SR models on perceptual quality and downstream task performance. We conduct experiments with 270 settings, covering 2 cross-platform SR tasks, 9 SR models, 3 downstream task models, and 5 downstream tasks for each SR task. The results show that improvements in traditional SR metrics often do not correlate with gains in task performance, and the correlations can be negative, indicating that these metrics provide limited guidance for selecting superior models for downstream tasks. This reveals the need to integrate downstream tasks into SR model development and evaluation.

Abstract

Explore similar work

Beyond Visual Fidelity: Benchmarking Super-Resolution Models for Large-Scale Remote Sensing Imagery via Downstream Task Integration

MapSR: Prompt-Driven Land Cover Map Super-Resolution via Vision Foundation Models

Simon-SR: Spatially Adaptive Modulation and Visual Prompt Adaptation for Text-Reinforced Super-Resolution