分割论文趋势 2023–2026

主要结论

• 分割论文从 2023 年的 251 篇增长到 2026 年的 265 篇，三年间扩大到约 1.1 倍。
• CVPR 是发分割论文最多的会议，2023–2026 年共 647 篇，占四个会议总量的 63%。
• 交叉方向里涨得最快的是多模态大模型：同时属于两个主题的论文从 2023 年的 36 篇增至 2026 年的 72 篇。
• 在「计算机视觉」的 5 个主题里，它的 2026 年论文数排第 5；同组增长最快的视频理解是 2.6 倍，分割是 1.1 倍。

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

• NeurIPS 2026A probabilistic model of visual segmentation explains early visual cortical dynamics
• ICML 2026SEMIR: Semantic Minor-Induced Representation Learning on Graphs for Visual Segmentation
• ICLR 2026ARINBEV: Bird's-Eye View Layout Estimation with Conditional Autoregressive Model
• CVPR 2026Annotation-Efficient Coreset Selection for Context-dependent Segmentation
• NeurIPS 2025$\epsilon$-Seg: Sparsely Supervised Semantic Segmentation of Microscopy Data
• ICML 2025GraphCL: Graph-based Clustering for Semi-Supervised Medical Image Segmentation
• ICLR 2025ADAPT: Attentive Self-Distillation and Dual-Decoder Prediction Fusion for Continual Panoptic Segmentation
• CVPR 20253D Dental Model Segmentation with Geometrical Boundary Preserving
• NeurIPS 2024Archaeoscape: Bringing Aerial Laser Scanning Archaeology to the Deep Learning Era
• ICML 2024BLO-SAM: Bi-level Optimization Based Finetuning of the Segment Anything Model for Overfitting-Preventing Semantic Segmentation
• ICLR 2024Diving Segmentation Model into Pixels
• CVPR 2024DiverGen: Improving Instance Segmentation by Learning Wider Data Distribution with More Diverse Generative Data
• NeurIPS 20233D Indoor Instance Segmentation in an Open-World
• ICML 2023CLUSTSEG: Clustering for Universal Segmentation
• ICLR 20233D Segmenter: 3D Transformer based Semantic Segmentation via 2D Panoramic Distillation
• CVPR 2023A Generalized Framework for Video Instance Segmentation
• NeurIPS 2026Adaptive Residual Quantization for Memory-Efficient Temporal Action Segmentation
• ICML 2026SCRWKV: Ultra-Compact Structure-Calibrated Vision-RWKV for Topological Crack Segmentation
• ICLR 2026CONSIGN: Conformal Segmentation Informed by Spatial Groupings via Decomposition
• CVPR 2026Better Stronger Faster Tackling the Trilemma in MLLM-based Segmentation with

全部分割论文

共 26 篇，第 1 / 1 页

1. ICLR 2025ADAPT: Attentive Self-Distillation and Dual-Decoder Prediction Fusion for Continual Panoptic Segmentation
2. ICLR 2025DECO: Unleashing the Potential of ConvNets for Query-based Detection and Segmentation
3. ICLR 2025Decoupled Finetuning for Domain Generalizable Semantic Segmentation
4. ICLR 2025DynAlign: Unsupervised Dynamic Taxonomy Alignment for Cross-Domain Segmentation
5. ICLR 2025Efficient Neuron Segmentation in Electron Microscopy by Affinity-Guided Queries
6. ICLR 2025GrabS: Generative Embodied Agent for 3D Object Segmentation without Scene Supervision
7. ICLR 2025Instance-dependent Early Stopping
8. ICLR 2025Interpreting the Second-Order Effects of Neurons in CLIP
9. ICLR 2025Learning Spatial-Semantic Features for Robust Video Object Segmentation
10. ICLR 2025Multimodality Helps Few-shot 3D Point Cloud Semantic Segmentation
11. ICLR 2025Near, far: Patch-ordering enhances vision foundation models' scene understanding
12. ICLR 2025Open-YOLO 3D: Towards Fast and Accurate Open-Vocabulary 3D Instance Segmentation
13. ICLR 2025Order-aware Interactive Segmentation
14. ICLR 2025Predictive Uncertainty Quantification for Bird's Eye View Segmentation: A Benchmark and Novel Loss Function
15. ICLR 2025Progressive Parameter Efficient Transfer Learning for Semantic Segmentation
16. ICLR 2025Progressive Token Length Scaling in Transformer Encoders for Efficient Universal Segmentation
17. ICLR 2025Revisit the Open Nature of Open Vocabulary Semantic Segmentation
18. ICLR 2025RMP-SAM: Towards Real-Time Multi-Purpose Segment Anything
19. ICLR 2025SAM-CP: Marrying SAM with Composable Prompts for Versatile Segmentation
20. ICLR 2025SegLLM: Multi-round Reasoning Segmentation with Large Language Models
21. ICLR 2025Segment Any 3D Object with Language
22. ICLR 2025Stable Segment Anything Model
23. ICLR 2025Topograph: An Efficient Graph-Based Framework for Strictly Topology Preserving Image Segmentation
24. ICLR 2025Towards Unbiased Learning in Semi-Supervised Semantic Segmentation
25. ICLR 2025Training-Free Dataset Pruning for Instance Segmentation
26. ICLR 2025UNIP: Rethinking Pre-trained Attention Patterns for Infrared Semantic Segmentation

常见问题

2023–2026 年四大会议一共有多少篇分割论文？

共 1,034 篇，按年份分别是 2023 年 251 篇、2024 年 269 篇、2025 年 249 篇、2026 年 265 篇。

NeurIPS 2026 有多少篇分割论文？

NeurIPS 2026 有 60 篇分割论文，上一届是 66 篇。

ICLR 2026 有多少篇分割论文？

ICLR 2026 有 26 篇分割论文，上一届是 26 篇。

ICML 2026 有多少篇分割论文？

ICML 2026 有 25 篇分割论文，上一届是 17 篇。

CVPR 2026 有多少篇分割论文？

CVPR 2026 有 154 篇分割论文，上一届是 140 篇。

哪个会议的分割论文最多？

CVPR，2023–2026 年共 647 篇。