2000 character limit reached
Regressor-Segmenter Mutual Prompt Learning for Crowd Counting
Published 4 Dec 2023 in cs.CV | (2312.01711v3)
Abstract: Crowd counting has achieved significant progress by training regressors to predict instance positions. In heavily crowded scenarios, however, regressors are challenged by uncontrollable annotation variance, which causes density map bias and context information inaccuracy. In this study, we propose mutual prompt learning (mPrompt), which leverages a regressor and a segmenter as guidance for each other, solving bias and inaccuracy caused by annotation variance while distinguishing foreground from background. In specific, mPrompt leverages point annotations to tune the segmenter and predict pseudo head masks in a way of point prompt learning. It then uses the predicted segmentation masks, which serve as spatial constraint, to rectify biased point annotations as context prompt learning. mPrompt defines a way of mutual information maximization from prompt learning, mitigating the impact of annotation variance while improving model accuracy. Experiments show that mPrompt significantly reduces the Mean Average Error (MAE), demonstrating the potential to be general framework for down-stream vision tasks.
- Variational information distillation for knowledge transfer. In CVPR, 2019.
- Leaf counting with deep convolutional and deconvolutional networks. In ICCV, 2017.
- Switching convolutional neural network for crowd counting. In CVPR, 2017.
- Adaptive dilated network with self-correction supervision for counting. In CVPR, 2020.
- Visual prompting via image inpainting. In NIPS, 2022.
- Language models are few-shot learners. In NIPS, 2020.
- Scale aggregation network for accurate and efficient crowd counting. In ECCV, 2018.
- Privacy preserving crowd monitoring: Counting people without people models or tracking. In CVPR, 2008.
- Variational attention: Propagating domain-specific knowledge for multi-domain learning in crowd counting. In ICCV, 2021.
- Rethinking spatial invariance of convolutional networks for object counting. In CVPR, 2022.
- Bert: Pre-training of deep bidirectional transformers for language understanding. In NAACL-HLT, 2019.
- Clrnet: A cross locality relation network for crowd counting in videos. TNNLS, 2022.
- Extremely overlapping vehicle counting. In IbPRIA, 2015.
- Steerer: Resolving scale variations for counting and localization via selective inheritance learning. In ICCV, 2023.
- Towards a unified view of parameter-efficient transfer learning. arXiv preprint arXiv:2110.04366, 2021.
- Parameter-efficient transfer learning for NLP. In ICML, 2019.
- Composition loss for counting, density map estimation and localization in dense crowds. In ECCV, 2018.
- Visual prompt tuning. In ECCV, 2022.
- Crowd counting and density estimation by trellis encoder-decoder networks. In CVPR, 2019a.
- Learning multi-level density maps for crowd counting. TNNLS, 2019b.
- Attention scaling for crowd counting. In CVPR, 2020a.
- How can we know what language models know? TACL, 2020b.
- Beyond counting: Comparisons of density maps for crowd analysis tasks—counting, detection, and tracking. TCSVT, 2018.
- Adam: A method for stochastic optimization. arXiv preprint, 2014.
- Segment anything. arXiv:2304.02643, 2023a.
- Segment anything. arXiv preprint arXiv:2304.02643, 2023b.
- Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25, 2012.
- Learning to count objects in images. In NIPS, 2010.
- Csrnet: Dilated convolutional neural networks for understanding the highly congested scenes. In CVPR, 2018.
- An end-to-end transformer model for crowd localization. In ECCV, 2022.
- Direct measure matching for crowd counting. IJCAI, 2021.
- Boosting crowd counting via multifaceted attention. In CVPR, 2022.
- Point-query quadtree for crowd counting, localization, and more. In ICCV, 2023.
- Crowd counting with deep structured scale integration network. In ICCV, 2019a.
- Adcrowdnet: An attention-injective deformable convolutional network for crowd understanding. In CVPR, 2019b.
- Pre-train, prompt, and predict: A systematic survey of prompting methods in natural language processing. ACM Computing Surveys, 2022.
- Context-aware crowd counting. In CVPR, 2019c.
- Swin transformer: Hierarchical vision transformer using shifted windows. In ICCV, 2021.
- Tasselnet: counting maize tassels in the wild via local counts regression network. Plant methods, 2017.
- Fantastically ordered prompts and where to find them: Overcoming few-shot prompt order sensitivity. arXiv preprint, 2021.
- Bayesian loss for crowd count estimation with point supervision. In ICCV, 2019.
- Understanding the impact of mistakes on background regions in crowd counting. In WCACV, 2021.
- Towards perspective-free object counting with deep learning. In ECCV, 2016.
- Training language models to follow instructions with human feedback. In NIPS, 2022.
- Language models are unsupervised multitask learners. OpenAI blog, 2019.
- Diffuse-denoise-count: Accurate crowd-counting with diffusion models. arXiv preprint, 2023.
- Crowd counting using multiple local features. In Digital Image Computing: Techniques and Applications, 2009.
- Revisiting perspective information for efficient crowd counting. In CVPR, 2019a.
- Counting with focus for free. In ICCV, 2019b.
- Crowd counting in the frequency domain. In CVPR, 2022.
- Very deep convolutional networks for large-scale image recognition. arXiv preprint, 2014.
- Rethinking counting and localization in crowds: A purely point-based framework. In ICCV, 2021a.
- To choose or to fuse? scale selection for crowd counting. In AAAI, 2021b.
- Modeling noisy annotations for crowd counting. In NIPS, 2020.
- A generalized loss function for crowd counting and localization. In CVPR, 2021.
- Distribution matching for crowd counting. In NIPS, 2020a.
- Uniformity in heterogeneity: Diving deep into count interval partition for crowd counting. In ICCV, 2021.
- Deep high-resolution representation learning for visual recognition. TPAMI, 2020b.
- Nwpu-crowd: A large-scale benchmark for crowd counting and localization. TPAMI, 2020c.
- Discrete-constrained regression for local counting models. arXiv preprint, 2022.
- Perspective-guided convolution networks for crowd counting. In ICCV, 2019.
- Crowd counting via perspective-guided fractional-dilation convolution. IEEE Transactions on Multimedia, 2021.
- Reverse perspective network for perspective-aware object counting. In CVPR, 2020.
- Relational attention network for crowd counting. In ICCV, 2019.
- Cross-scene crowd counting via deep convolutional neural networks. In CVPR, 2015.
- Single-image crowd counting via multi-column convolutional neural network. In CVPR, 2016.
- Leveraging heterogeneous auxiliary tasks to assist crowd counting. In CVPR, 2019.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.
