mlsanityData to 14 Oct 2026

Real-Time Crowd Counting for Embedded Systems with Lightweight Architecture

arXiv 2510.13250 ↗First posted 15 Oct 2025cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Crowd counting is a task of estimating the number of the crowd through images, which is extremely valuable in the fields of intelligent security, urban planning, public safety management, and so on. However, the existing counting methods have some problems in practical application on embedded systems for these fields, such as excessive model parameters, abundant complex calculations, etc. The practical application of embedded systems requires the model to be real-time, which means that the model is fast enough. Considering the aforementioned problems, we design a super real-time model with a stem-encoder-decoder structure for crowd counting tasks, which achieves the fastest inference compared with state-of-the-arts. Firstly, large convolution kernels in the stem network are used to enlarge the receptive field, which effectively extracts detailed head information. Then, in the encoder part, we use conditional channel weighting and multi-branch local fusion block to merge multi-scale features with low computational consumption. This part is crucial to the super real-time performance of the model. Finally, the feature pyramid networks are added to the top of the encoder to alleviate its incomplete fusion problems. Experiments on three benchmarks show that our network is suitable for super real-time crowd counting on embedded systems, ensuring competitive accuracy. At the same time, the proposed network reasoning speed is the fastest. Specifically, the proposed network achieves 381.7 FPS on NVIDIA GTX 1080Ti and 71.9 FPS on NVIDIA Jetson TX1.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 37 of its 73 references that are in the map. First posted on arXiv 15 Oct 2025.

Builds on

37 of 73 references.

References · 37

  1. 2016Feature Pyramid Networks for Object Detectioncited by 5,036
  2. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  3. 2017MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applicationscited by 3,845
  4. 2017ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devicescited by 1,200
  5. 2018ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Designcited by 1,136
  6. 2016SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model sizecited by 1,060
  7. 2018CSRNet: Dilated Convolutional Neural Networks for Understanding the Highly Congested Scenescited by 267
  8. 2018Composition Loss for Counting, Density Map Estimation and Localization in Dense Crowdscited by 186
  9. 2020NWPU-Crowd: A Large-Scale Benchmark for Crowd Counting and Localizationcited by 151
  10. 2019Learning from Synthetic Data for Crowd Counting in the Wildcited by 148
  11. 2021Lite-HRNet: A Lightweight High-Resolution Networkcited by 91
  12. 2025AI Flow: Perspectives, Scenarios, and Approachescited by 44
25 more
  1. 2017CNN-based Cascaded Multi-task Learning of High-level Prior and Density Estimation for Crowd Countingcited by 41
  2. 2022Positive-incentive Noisecited by 39
  3. 2023Variational Positive-incentive Noise: How Noise Benefits Modelscited by 38
  4. 2021Reconstructive Sequence-Graph Network for Video Summarizationcited by 34
  5. 2024Enhance Vision-Language Alignment with Noisecited by 34
  6. 2019PCC Net: Perspective Crowd Counting via Spatial Convolutional Networkcited by 33
  7. 2017Spatiotemporal Modeling for Crowd Counting in Videoscited by 27
  8. 2024Data Augmentation of Contrastive Learning is Estimating Positive-incentive Noisecited by 25
  9. 2019C^3 Framework: An Open-source PyTorch Code for Crowd Countingcited by 18
  10. 2022DR.VIC: Decomposition and Reasoning for Video Individual Countingcited by 18
  11. 2025Learn Beneficial Noise as Graph Augmentationcited by 15
  12. 2025WebUIBench: A Comprehensive Benchmark for Evaluating Multimodal Large Language Models in WebUI-to-Codecited by 15
  13. 2018Top-Down Feedback for Crowd Counting Convolutional Neural Networkcited by 14
  14. 2025Mixture of Noise for Pre-Trained Model-Based Class-Incremental Learningcited by 14
  15. 2020Efficient Crowd Counting via Structured Knowledge Transfercited by 13
  16. 2024A Fixed-Point Approach to Unified Prompt-Based Countingcited by 10
  17. 2023Combating Data Imbalances in Federated Semi-supervised Learning with Dual Regulatorscited by 9
  18. 2025LLMs Caught in the Crossfire: Malware Requests and Jailbreak Challengescited by 9
  19. 2020A Real-Time Deep Network for Crowd Countingcited by 7
  20. 2025Scale Efficient Training for Large Datasetscited by 7
  21. 2024Diffusion-based Data Augmentation for Object Counting Problemscited by 5
  22. 2024Robust Zero-Shot Crowd Counting and Localization With Adaptive Resolution SAMcited by 4
  23. 2024Embedding Generalized Semantic Knowledge into Few-Shot Remote Sensing Segmentationcited by 2
  24. 2021Scale-Aware Network with Regional and Semantic Attentions for Crowd Counting under Cluttered Backgroundcited by 1
  25. 2025Density-based Object Detection in Crowded Scenescited by 1

Built on by

Not cited yet.