Adaptive NMS: Refining Pedestrian Detection in a Crowd

By Songtao Liu, Di Huang, Yunhong Wang. 2019

340 citations recorded by the source corpus.

Pedestrian detection in a crowd is a very challenging issue. This paper addresses this problem by a novel Non-Maximum Suppression (NMS) algorithm to better refine the bounding boxes given by detectors. The contributions are threefold: (1) we propose adaptive-NMS, which applies a dynamic suppression threshold to an instance, according to the target density; (2) we design an efficient subnetwork to learn density scores, which can be conveniently embedded into both the single-stage and two-stage detectors; and (3) we achieve state of the art results on the CityPersons and CrowdHuman benchmarks.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Chunhua Shen
2. Alan L. Yuille
3. Philip H. S. Torr
4. Xinggang Wang
5. Bernt Schiele
6. Dacheng Tao
7. Trevor Darrell
8. Quoc V. Le

In Computer Vision and Pattern Recognition

1. Luc Van Gool
2. Chunhua Shen
3. Alan L. Yuille
4. Ming-Hsuan Yang
5. Xiaogang Wang
6. Dacheng Tao
7. Chen Change Loy
8. Andrew Zisserman

In Signal Processing

1. Xiang Bai
2. Chunhua Shen
3. Junsong Yuan
4. Serge J. Belongie
5. Xinggang Wang
6. X. Zhang
7. Xuangeng Chu
8. Anlin Zheng

Citations in the corpus

• Occluded Video Instance Segmentation: A Benchmark (2021, methodology): “In addition, some learnable NMS methods [29,53] have also been proposed, and many new methods [9,20] based on set prediction even do not need NMS post-processing.”
• Distance-IoU Loss: Faster and Better Learning for Bounding Box Regression (2019, background): “More-over, the DIoU-NMS is very flexible to be integrated into any object detection pipeline with only a few lines of code.”
• SOLOv2: Dynamic and Fast Instance Segmentation (2020, background): “Some recent works [1, 26, 13, 3, 2] are proposed to improve the traditional NMS.”
• Enhancing Geometric Factors in Model Learning and Inference for Object Detection and Instance Segmentation (2020, background): “Recently, Adaptive NMS [51] and Softer-NMS [52] are proposed to respectively study proper threshold and weighted average strategies.”
• OTA: Optimal Transport Assignment for Object Detection (2021, background): “Object detection in crowded scenarios has raised more and more attention [24, 15, 9, 10].”
• Occluded Video Instance Segmentation: Dataset and ICCV 2021 Challenge (2021, background): “[25] introduces adaptive-NMS which adaptively increases the NMS threshold in crowd scenes.”
• Detection in Crowded Scenes: One Proposal, Multiple Predictions (2020, methodology): “Other works such as [23, 17] propose to predict different NMS thresholds for different bounding boxes.”
• Temporal-Context Enhanced Detection of Heavily Occluded Pedestrians (2020, background): “With the renaissance of convolutional neural networks, many deep learning based methods on pedestrian detection [27, 28, 10, 26, 24, 23, 19, 17, 25, 18, 48, 36] significantly outperform the hand-crafted feature based methods [55, 61, 21, 14].”
• BlockCopy: High-Resolution Video Processing with Block-Sparse Feature Propagation and Online Policies (2021, neutral): “270 s AdaptiveNMS (CVPR2019) [23] VGG-16 11.”