Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
143 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
46 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Curriculum for Crowd Counting -- Is it Worthy? (2401.07586v1)

Published 15 Jan 2024 in cs.CV and cs.AI

Abstract: Recent advances in deep learning techniques have achieved remarkable performance in several computer vision problems. A notably intuitive technique called Curriculum Learning (CL) has been introduced recently for training deep learning models. Surprisingly, curriculum learning achieves significantly improved results in some tasks but marginal or no improvement in others. Hence, there is still a debate about its adoption as a standard method to train supervised learning models. In this work, we investigate the impact of curriculum learning in crowd counting using the density estimation method. We performed detailed investigations by conducting 112 experiments using six different CL settings using eight different crowd models. Our experiments show that curriculum learning improves the model learning performance and shortens the convergence time.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (48)
  1. Curriculum learning. In Proceedings of the 26th Annual International Conference on Machine Learning, ICML ’09, page 41–48, New York, NY, USA. Association for Computing Machinery.
  2. Crowdnet: A deep convolutional network for dense crowd counting. Proceedings of the 24th ACM international conference on Multimedia.
  3. Scale aggregation network for accurate and efficient crowd counting. In ECCV.
  4. Bayesian poisson regression for crowd counting. In 2009 IEEE 12th International Conference on Computer Vision, pages 545–551.
  5. Feature mining for localised crowd counting. In BMVC.
  6. Webly supervised learning of convolutional networks. 2015 IEEE International Conference on Computer Vision (ICCV), pages 1431–1439.
  7. A survey of crowd counting and density estimation based on convolutional neural network. Neurocomputing, 472:224–251.
  8. Locality-constrained spatial transformer network for video crowd counting. In 2019 IEEE International Conference on Multimedia and Expo (ICME), pages 814–819.
  9. Mobilecount: An efficient encoder-decoder framework for real-time crowd counting. In Pattern Recognition and Computer Vision: Second Chinese Conference, PRCV 2019, Xi’an, China, November 8–11, 2019, Proceedings, Part II, page 582–595. Springer-Verlag.
  10. Advances in convolution neural networks based crowd counting and density estimation. Big Data Cogn. Comput., 5:50.
  11. A unified multi-task learning framework of real-time drone supervision for crowd counting. ArXiv, abs/2202.03843.
  12. Deep residual learning for image recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 770–778.
  13. Composition loss for counting, density map estimation and localization in dense crowds. ArXiv, abs/1808.01050.
  14. How hard can it be? estimating the difficulty of visual search in an image. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 2157–2166.
  15. Crowd counting and density estimation by trellis encoder-decoder networks. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6126–6135.
  16. Clip: Train faster with less data. In 2023 IEEE International Conference on Big Data and Smart Computing (BigComp), pages 34–39. IEEE.
  17. Revisiting Crowd Counting: State-of-the-art, Trends, and Future Perspectives. arXiv preprint arXiv:2209.07271.
  18. LCDnet: a lightweight crowd density estimation model for real-time video surveillance. Journal of Real-Time Image Processing, 20:29.
  19. Curriculum learning and minibatch bucketing in neural machine translation. In RANLP.
  20. Krizhevsky, A. (2009). Learning multiple layers of features from tiny images. Technical report.
  21. Estimating the number of people in crowded scenes by mid based foreground segmentation and head-shoulder detection. In 2008 19th International Conference on Pattern Recognition, pages 1–4.
  22. Multiple instance curriculum learning for weakly supervised object detection. ArXiv, abs/1711.09191.
  23. Learning error-driven curriculum for crowd counting. 2020 25th International Conference on Pattern Recognition (ICPR), pages 843–849.
  24. Csrnet: Dilated convolutional neural networks for understanding the highly congested scenes. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1091–1100.
  25. Bayesian loss for crowd count estimation with point supervision. In 2019 IEEE/CVF International Conference on Computer Vision (ICCV), pages 6141–6150, Los Alamitos, CA, USA.
  26. Curriculum learning for reinforcement learning domains: A framework and survey. J. Mach. Learn. Res., 21:181:1–181:50.
  27. Rgb-t crowd counting from drone: A benchmark and mmccn network. In ACCV.
  28. Competence-based curriculum learning for neural machine translation. ArXiv, abs/1903.09848.
  29. Pixel-wise crowd understanding via synthetic data. International Journal of Computer Vision, 129:225–245.
  30. Mobilenetv2: Inverted residuals and linear bottlenecks. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 4510–4520.
  31. Self paced deep learning for weakly supervised object detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 41:712–725.
  32. Weakly supervised object localization using size estimates. ArXiv, abs/1608.04314.
  33. A real-time deep network for crowd counting. ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 2328–2332.
  34. Very deep convolutional networks for large-scale image recognition. In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings.
  35. Cnn-based cascaded multi-task learning of high-level prior and density estimation for crowd counting. In 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), pages 1–6, Los Alamitos, CA, USA. IEEE Computer Society.
  36. To choose or to fuse? scale selection for crowd counting. In AAAI.
  37. Going deeper with convolutions. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 1–9.
  38. Attention-guided curriculum learning for weakly supervised classification and localization of thoracic diseases on chest radiographs. ArXiv, abs/1807.07532.
  39. Counting people by clustering person detector outputs. In 2014 11th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), pages 313–318.
  40. Crowd counting via segmentation guided attention networks and curriculum loss. IEEE Transactions on Intelligent Transportation Systems.
  41. Density-aware curriculum learning for crowd counting. IEEE Transactions on Cybernetics, 52:4675–4687.
  42. Dynamically composing domain-data selection with clean-data selection by “co-curricular learning” for neural machine translation. In ACL.
  43. A survey on curriculum learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 44(9):4555–4576.
  44. When do curricula work? ArXiv, abs/2012.03107.
  45. Weakly-supervised crowd counting learns from sorting rather than locations. In Computer Vision – ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part VIII, page 1–17, Berlin, Heidelberg. Springer-Verlag.
  46. Multi-scale convolutional neural networks for crowd counting. 2017 IEEE International Conference on Image Processing (ICIP), pages 465–469.
  47. Cross-scene crowd counting via deep convolutional neural networks. 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 833–841.
  48. Single-image crowd counting via multi-column convolutional neural network. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 589–597.
Citations (1)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now