NeurIPS 2025 accepted papers on Detection and segmentation
113 of the 5,812 papers accepted at NeurIPS 2025 (1.9%) are on Detection and segmentation, compared with 2.8% at NeurIPS 2024. They fall into 5 topics.
See NeurIPS 2025 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Open-set and few-shot detection · 62 papers
- PanTS: The Pancreatic Tumor Segmentation Dataset (7 citations)
- Extremely Simple Multimodal Outlier Synthesis for Out-of-Distribution Detection and Segmentation (4 citations)
- OmniTry: Virtual Try-On Anything without Masks (4 citations)
- SANSA: Unleashing the Hidden Semantics in SAM2 for Few-Shot Segmentation (4 citations)
- InstructSAM: A Training-free Framework for Instruction-Oriented Remote Sensing Object Recognition (3 citations)
- ArchCAD-400K: A Large-Scale CAD drawings Dataset and New Baseline for Panoptic Symbol Spotting (2 citations)
- Mars-Bench: A Benchmark for Evaluating Foundation Models for Mars Science Tasks (2 citations)
- MetaSlot: Break Through the Fixed Number of Slots in Object-Centric Learning (2 citations)
- Object-Centric Representation Learning for Enhanced 3D Semantic Scene Graph Prediction (2 citations)
- RESAnything: Attribute Prompting for Arbitrary Referring Segmentation (2 citations)
- Test-Time Adaptive Object Detection with Foundation Model (2 citations)
- COS3D: Collaborative Open-Vocabulary 3D Segmentation (1 citation)
and 50 more in the interactive map.
CNN features · 29 papers
- YOLOv12: Attention-Centric Real-Time Object Detectors (637 citations)
- DAMamba: Vision State Space Model with Dynamic Adaptive Scan (3 citations)
- Delving into Cascaded Instability: A Lipschitz Continuity View on Image Restoration and Object Detection Synergy (2 citations)
- UniSite: The First Cross-Structure Dataset and Learning Framework for End-to-End Ligand Binding Site Detection (2 citations)
- End-to-End Low-Light Enhancement for Object Detection with Learned Metadata from RAWs (1 citation)
- FlexEvent: Towards Flexible Event-Frame Object Detection at Varying Operational Frequencies (1 citation)
- FRBNet: Revisiting Low-Light Vision through Frequency-Domain Radial Basis Network (1 citation)
- MixPrompt: Efficient Mixed Prompting for Multimodal Semantic Segmentation (1 citation)
- Real-Time Scene-Adaptive Tone Mapping for High-Dynamic Range Object Detection (1 citation)
- Rethinking Scale-Aware Temporal Encoding for Event-based Object Detection (1 citation)
- DeepASA: An Object-Oriented Multi-Purpose Network for Auditory Scene Analysis
- Diffusion-Driven Two-Stage Active Learning for Low-Budget Semantic Segmentation
and 17 more in the interactive map.
3D detection and driving · 10 papers
- OpenAD: Open-World Autonomous Driving Benchmark for 3D Object Detection (8 citations)
- Availability-aware Sensor Fusion via Unified Canonical Space (3 citations)
- OpenBox: Annotate Any Bounding Boxes in 3D (1 citation)
- CodeMerge: Codebook-Guided Model Merging for Robust Test-Time Adaptation in Autonomous Driving
- No Object Is an Island: Enhancing 3D Semantic Segmentation Generalization with Diffusion Models
- RayFusion: Ray Fusion Enhanced Collaborative Visual Perception
- SQS: Enhancing Sparse Perception Models via Query-based Splatting in Autonomous Driving
- T-norm Selection for Object Detection in Autonomous Driving with Logical Constraints
- Towards 3D Objectness Learning in an Open World
- VoxDet: Rethinking 3D Semantic Scene Completion as Dense Object Detection
Tracking · 10 papers
- Tracking and Understanding Object Transformations (2 citations)
- DOVTrack: Data-Efficient Open-Vocabulary Tracking (1 citation)
- Dual-Path Temporal Decoder for End-to-End Multi-Object Tracking (1 citation)
- MMOT: The First Challenging Benchmark for Drone-based Multispectral Multi-Object Tracking (1 citation)
- InFlux: A Benchmark for Self-Calibration of Dynamic Intrinsics of Video Cameras
- Is This Tracker On? A Benchmark Protocol for Dynamic Tracking
- Lattice Boltzmann Model for Learning Real-World Pixel Dynamicity
- STAR: Spatial-Temporal Tracklet Matching for Multi-Object Tracking
- SynCL: A Synergistic Training Strategy with Instance-Aware Contrastive Learning for End-to-End Multi-Camera 3D Tracking
- Video Diffusion Models Excel at Tracking Similar-Looking Objects Without Supervision