mlsanityData to 14 Oct 2026

Learning to Count without Annotations

arXiv 2307.08727 ↗First posted 17 Jul 2023cs.CVcited by 5

Not on the map. Why?

Abstract (arXiv)

While recent supervised methods for reference-based object counting continue to improve the performance on benchmark datasets, they have to rely on small datasets due to the cost associated with manually annotating dozens of objects in images. We propose UnCounTR, a model that can learn this task without requiring any manual annotations. To this end, we construct "Self-Collages", images with various pasted objects as training samples, that provide a rich learning signal covering arbitrary object types and counts. Our method builds on existing unsupervised representations and segmentation techniques to successfully demonstrate for the first time the ability of reference-based counting without manual supervision. Our experiments show that our method not only outperforms simple baselines and generic models such as FasterRCNN and DETR, but also matches the performance of supervised counting models in some domains.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Builds on 56 of its 76 references that are in the map. First posted on arXiv 17 Jul 2023.

Builds on

56 of 76 references.

References · 56

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  5. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 12,049
  6. 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,639
  7. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,498
  8. 2023Segment Anythingcited by 8,234
  9. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  10. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  11. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,513
  12. 2020End-to-End Object Detection with Transformerscited by 7,414
38 more
  1. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  2. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  3. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,750
  4. 2019Analyzing and Improving the Image Quality of StyleGANcited by 3,588
  5. 2016Context Encoders: Feature Learning by Inpaintingcited by 1,428
  6. 2016Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzlescited by 1,382
  7. 2015Unsupervised Visual Representation Learning by Context Predictioncited by 1,066
  8. 2018Deep Clustering for Unsupervised Learning of Visual Featurescited by 1,028
  9. 2016Sequence-Level Knowledge Distillationcited by 801
  10. 2022Detecting Twenty-thousand Classes using Image-level Supervisioncited by 614
  11. 2020Simple Copy-Paste is a Strong Data Augmentation Method for Instance Segmentationcited by 596
  12. 2019Self-labelling via simultaneous clustering and representation learningcited by 528
  13. 2020SCAN: Learning to Classify Images without Labelscited by 352
  14. 2020Self-Supervised MultiModal Versatile Networkscited by 276
  15. 2021Localizing Objects with Self-Supervised Transformers and no Labelscited by 230
  16. 2019Video Representation Learning by Dense Predictive Codingcited by 175
  17. 2022Correct-N-Contrast: A Contrastive Approach for Improving Robustness to Spurious Correlationscited by 172
  18. 2017Drone-based Object Counting by Spatially Regularized Regional Proposal Networkcited by 156
  19. 2021Learning To Count Everythingcited by 150
  20. 2022Deep Spectral Methods: A Surprisingly Strong Baseline for Unsupervised Semantic Segmentation and Localizationcited by 150
  21. 2017Representation Learning by Learning to Countcited by 120
  22. 2018Modeling Visual Context is Key to Augmenting Object Detection Datasetscited by 106
  23. 2021Learning to See by Looking at Noisecited by 92
  24. 2022Self-Supervised Learning of Object Parts for Semantic Segmentationcited by 90
  25. 2016A Large Contextual Dataset for Classification, Detection and Counting of Cars with Deep Learningcited by 86
  26. 2018Class-Agnostic Countingcited by 85
  27. 2019ClusterFit: Improving Generalization of Visual Representationscited by 82
  28. 2020Hard negative examples are hard, but usefulcited by 82
  29. 2022CounTR: Transformer-based Generalised Visual Countingcited by 79
  30. 2022A Low-Shot Object Counting Network With Iterative Prototype Adaptationcited by 65
  31. 2022Few-shot Object Counting with Similarity-Aware Feature Enhancementcited by 62
  32. 2018Leveraging Unlabeled Data for Crowd Counting by Learning to Rankcited by 57
  33. 2022Unsupervised Salient Object Detection with Spectral Cluster Votingcited by 56
  34. 2022Unsupervised Semantic Segmentation with Self-supervised Object-centric Representationscited by 54
  35. 2023CrowdCLIP: Unsupervised Crowd Counting via Vision-Language Modelcited by 47
  36. 2021Robust Contrastive Learning Using Negative Samples with Diminished Semanticscited by 43
  37. 2022Learning to Count Anything: Reference-less Class-agnostic Counting with Weak Supervisioncited by 41
  38. 2023Training-free Object Counting with Promptscited by 34

50 most cited of 56 All 56 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 2.

5 papers.

  1. 2025A Survey on Class-Agnostic Counting: Advancements from Reference-Based to Open-World Text-Guided Approachescited by 7
  2. 2025CountingDINO: A Training-free Pipeline for Class-Agnostic Counting using Unsupervised Backbonescited by 6
  3. 2025Consistent-Point: Consistent Pseudo-Points for Semi-Supervised Crowd Counting and Localizationcited by 1
  4. 2026Object Counting Across Modalities: Taxonomies, Benchmarks, Applications, and Open Challengescited by 0
  5. 2026Bottom-up Modeling of Repeated Elements via Single Image Analysis-by-Synthesiscited by 0