# Work-Efficient Parallel Non-Maximum Suppression Kernels

arXiv: https://arxiv.org/abs/2502.00535 · HTML: https://mlsanity.com/paper/2502.00535/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 13 of its 36 references that are in the map.
- First posted on arXiv 1 Feb 2025.

## Builds on (most cited first, 13 of 13)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [SSD: Single Shot MultiBox Detector](https://mlsanity.com/paper/1512.02325/) (2015)
- [Rich feature hierarchies for accurate object detection and semantic segmentation](https://mlsanity.com/paper/1311.2524/) (2013)
- [Joint Face Detection and Alignment using Multi-task Cascaded Convolutional Networks](https://mlsanity.com/paper/1604.02878/) (2016)
- [Soft-NMS -- Improving Object Detection With One Line of Code](https://mlsanity.com/paper/1704.04503/) (2017)
- [CrowdHuman: A Benchmark for Detecting Human in a Crowd](https://mlsanity.com/paper/1805.00123/) (2018)
- [Learning non-maximum suppression](https://mlsanity.com/paper/1705.02950/) (2017)
- [NVIDIA Tensor Core Programmability, Performance & Precision](https://mlsanity.com/paper/1803.04014/) (2018)
- [Adaptive NMS: Refining Pedestrian Detection in a Crowd](https://mlsanity.com/paper/1904.03629/) (2019)
- [FaceBoxes: A CPU Real-time Face Detector with High Accuracy](https://mlsanity.com/paper/1708.05234/) (2017)
- [An Empirical Evaluation of Deep Learning on Highway Driving](https://mlsanity.com/paper/1504.01716/) (2015)
- [FeatureNMS: Non-Maximum Suppression by Learning Feature Embeddings](https://mlsanity.com/paper/2002.07662/) (2020)
- [Multi-view Face Detection Using Deep Convolutional Neural Networks](https://mlsanity.com/paper/1502.02766/) (2015)

## Built on by (most cited first, 1 of 1)

- [Differentiable NMS via Sinkhorn Matching for End-to-End Fabric Defect Detection](https://mlsanity.com/paper/2505.07040/) (2025)

## Abstract (arXiv)

In the context of object detection, sliding-window classifiers and single-shot Convolutional Neural Network (CNN) meta-architectures typically yield multiple overlapping candidate windows with similar high scores around the true location of a particular object. Non-Maximum Suppression (NMS) is the process of selecting a single representative candidate within this cluster of detections, so as to obtain a unique detection per object appearing on a given picture. In this paper, we present a highly scalable NMS algorithm for embedded GPU architectures that is designed from scratch to handle workloads featuring thousands of simultaneous detections on a given picture. Our kernels are directly applicable to other sequential NMS algorithms such as FeatureNMS, Soft-NMS or AdaptiveNMS that share the inner workings of the classic greedy NMS method. The obtained performance results show that our parallel NMS algorithm is capable of clustering 1024 simultaneous detected objects per frame in roughly 1 ms on both NVIDIA Tegra X1 and NVIDIA Tegra X2 on-die GPUs, while taking 2 ms on NVIDIA Tegra K1. Furthermore, our proposed parallel greedy NMS algorithm yields a 14x-40x speed up when compared to state-of-the-art NMS methods that require learning a CNN from annotated data.