YOLO26: An Analysis of NMS-Free End to End Framework for Real-Time Object Detection

2026-01-19 · Sudip Chakrabarty arxiv

The ``You Only Look Once'' (YOLO) framework has long served as a standard for real-time object detection, though traditional iterations have utilized Non-Maximum Suppression (NMS) post-processing, which introduces specific latency and hyperparameter variables. This paper presents a comprehensive architectural analysis of YOLO26, a model that shifts toward a native end-to-end learning strategy by eliminating NMS. This study examines the core mechanisms driving this framework: the MuSGD optimizer for backbone stabilization, Small-Target-Aware Label Assignment (STAL), and ProgLoss for dynamic supervision. To contextualize its performance, this article reviews exhaustive benchmark data from the COCO \texttt{val2017} leaderboard. This evaluation provides an objective comparison of YOLO26 across various model scales (Nano to Extra-Large) against both prior CNN lineages and contemporary Transformer-based architectures (e.g., RT-DETR, DEIM, RF-DETR), detailing the observed speed-accuracy trade-offs and parameter requirements without asserting a singular optimal model. Additionally, the analysis covers the framework's unified multi-task capabilities, including the YOLOE-26 open-vocabulary module for promptable detection. Ultimately, this paper serves to document how decoupling representation learning from heuristic post-processing impacts the "Export Gap" and deterministic latency in modern edge-based computer vision deployments.

📄 PDF Abstract BibTeX arXiv:2601.12882

Code (0)

등록된 구현이 없습니다.

Tasks

Real-Time Object Detection Representation Learning

YOLO26 vs. YOLOv8: A Comprehensive Architectural Benchmark of Next-Generation Real-Time Object Detection Models

2026-05-24 · Chidera G. Oguine, Kanyifeechukwu J. Oguine, Obiozor M. Oguine, Ozioma C. Oguine arxiv

This paper presents a rigorous empirical evaluation of Ultralytics YOLO26 against the YOLOv8 baseline, offering an independent real-world stress test of NMS-free architectures on non-COCO distributions. Engineered for ed…

Real-Time Object Detection

YOLO26: Key Architectural Enhancements and Performance Benchmarking for Real-Time Object Detection

2025-09-29 · Ranjan Sapkota, Rahul Harsha Cheppally, Ajay Sharda, Manoj Karkee arxiv

This study presents a comprehensive analysis of Ultralytics YOLO26(also called as YOLOv26), highlighting its key architectural enhancements and performance benchmarking for real-time object detection. YOLO26, released in…

Real-Time Object Detection Instance Segmentation

An Analysis of Layer-Freezing Strategies for Enhanced Transfer Learning in YOLO Architectures

2025-09-05 · Andrzej D. Dobrzycki, Ana M. Bernardos, José R. Casar arxiv

The You Only Look Once (YOLO) architecture is crucial for real-time object detection. However, deploying it in resource-constrained environments such as unmanned aerial vehicles (UAVs) requires efficient transfer learnin…

Real-Time Object Detection Transfer Learning

YOLOE-26: Integrating YOLO26 with YOLOE for Real-Time Open-Vocabulary Instance Segmentation

2026-01-29 · Ranjan Sapkota, Manoj Karkee arxiv

This paper presents YOLOE-26, a unified framework that integrates the deployment-optimized YOLO26(or YOLOv26) architecture with the open-vocabulary learning paradigm of YOLOE for real-time open-vocabulary instance segmen…

Instance Segmentation

SDD-YOLO: A Small-Target Detection Framework for Ground-to-Air Anti-UAV Surveillance with Edge-Efficient Deployment

2026-03-26 · Pengyu Chen, Haotian Sa, Yiwei Hu, Yuhan Cheng 외 arxiv

Detecting small unmanned aerial vehicles (UAVs) from a ground-to-air (G2A) perspective presents significant challenges, including extremely low pixel occupancy, cluttered aerial backgrounds, and strict real-time constrai…

Object Detection