AI Paper Atlas

Research map / Benchmarks / RESISC45

RESISC45 leaderboard

10 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on RESISC45. Best accuracy: 97.48 (SatMAE++, CVPR 2024).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

RESISC45 · accuracy

10 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
197.48SatMAE++
ViT-Large
Rethinking Transformers Pre-training for Multi-Spectral Satellite Imagery
CVPR 2024 · Table 6
Human and embodied modeling
294.75LoRA-Pro
CLIP-ViT-B/16
LoRA-Pro: Are Low-Rank Adapters Properly Optimized?
ICLR 2025 · Image classification results t
Pruning and efficient tuning
394.57Auto-Switch
CLIP ViT-B/32
Less is More: Efficient Model Merging with Binary Task Switch
CVPR 2025 · Table 2
Pruning and efficient tuning
494.45GeoLink
ViT-L
GeoLink: Empowering Remote Sensing Foundation Model with OpenStreetMap Data
NeurIPS 2025 · Table: classification benchmar
Human and embodied modeling
592.6V2C-CBM
CLIP ViT-L/14
V2C-CBM: Building Concept Bottlenecks with Vision-to-Concept Tokenizer
AAAI 2025 · Table 1
Human and embodied modeling
690.98VTVC (sample-specific, probabilistic)
ViT-B/32
Variational Task Vector Composition
NeurIPS 2025 · Table: Performance comparison
Uncertainty and Gaussian processes
788.82CAP (Ours)
CLIP ViT-B/32
Handling Imbalanced Pseudolabels for Vision-Language Models with Concept Alignment and Confusion-Aware Calibrated Margin
ICML 2025 · Table 1
Multimodal LLMs
885.77Mom/Inv CNN (ours)On Making $SE(2)$-Invariant Networks Optimal
NeurIPS 2026 · Table 1
CNN design
980.76CURE
MobileCLIP2-B (device) + EVA-CLIP-18B (server)
CURE: Cumulative Knowledge Reuse for Efficient Device-Server Hybrid Inference in Vision-Language Models
ECCV 2026 · Table 1
Robot action and navigation
1079.7SVFT Banded d=2
ViT-L
SVFT: Parameter-Efficient Fine-Tuning with Singular Vectors
NeurIPS 2024 · Table: Image Classification
Pruning and efficient tuning

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.