# CountFormer: A Transformer Framework for Learning Visual Repetition and Structure in Class-Agnostic Object Counting

arXiv: https://arxiv.org/abs/2510.23785 · HTML: https://mlsanity.com/paper/2510.23785/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 19 references that are in the map.
- First posted on arXiv 27 Oct 2025.

## Builds on (most cited first, 15 of 15)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Teaching CLIP to Count to Ten](https://mlsanity.com/paper/2302.12066/) (2023)
- [Learning To Count Everything](https://mlsanity.com/paper/2104.08391/) (2021)
- [A Large Contextual Dataset for Classification, Detection and Counting of Cars with Deep Learning](https://mlsanity.com/paper/1609.04453/) (2016)
- [Class-Agnostic Counting](https://mlsanity.com/paper/1811.00472/) (2018)
- [CLIP-Count: Towards Text-Guided Zero-Shot Object Counting](https://mlsanity.com/paper/2305.07304/) (2023)
- [CounTR: Transformer-based Generalised Visual Counting](https://mlsanity.com/paper/2208.13721/) (2022)
- [Zero-shot Object Counting](https://mlsanity.com/paper/2303.02001/) (2023)
- [Learning to Count Anything: Reference-less Class-agnostic Counting with Weak Supervision](https://mlsanity.com/paper/2205.10203/) (2022)
- [Open-world Text-specified Object Counting](https://mlsanity.com/paper/2306.01851/) (2023)
- [Exemplar Free Class Agnostic Counting](https://mlsanity.com/paper/2205.14212/) (2022)
- [No Token Left Behind: Explainability-Aided Image Classification and Generation](https://mlsanity.com/paper/2204.04908/) (2022)

## Abstract (arXiv)

Humans can often count unfamiliar objects by observing visual repetition and composition, rather than relying only on object categories. However, many exemplar-free counting models struggle in such situations and may overcount when objects contain symmetric components, repeated substructures, or partial occlusion. We introduce CountFormer, a controlled adaptation of a density-regression framework inspired by CounTR, where the image encoder is replaced with the self-supervised vision foundation model DINOv2. The resulting transformer features are combined with explicit two-dimensional positional embeddings and decoded by a lightweight convolutional network to produce a density map whose integral gives the final count. Our goal is not to propose a new counting architecture, but to study whether foundation-based representations improve structural consistency under a strictly exemplar-free setting. On FSC-147, CountFormer achieves competitive performance under the official benchmark (MAE 19.06, RMSE 118.45). Qualitative analysis suggests fewer part-level overcounting errors for some structurally complex objects, while overall error remains broadly consistent with prior approaches. Sensitivity analysis shows that evaluation metrics are strongly affected by a small number of extreme high-density scenes. Overall, the results highlight the role of representation quality in exemplar-free object counting.