Research map / Detection and segmentation / Open-set and few-shot detection
Zero-shot Object Counting with Good Exemplars
Huilin Zhu, Jingling Yuan, Zhengwei Yang, Yu Guo, Xian Zhong, Zheng Wang, Shengfeng He
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add single-object exemplar filter to Grounding DINO (ablation) | Grounding DINO · CARPK | MAE: 29.72 -> 18.54 | Improved |
| add single-object filter without training the counter (ablation) | Grounding DINO exemplars · FSC-147 test | MAE: 54.48 -> 54.27 | No clear effect |
| add density loss training (ablation) | Grounding DINO + filter · FSC-147 test | MAE: 54.27 -> 18.93 | Improved |
| add contrastive loss with negative exemplars (ablation) | VA-Count · FSC-147 | MAE (val/test) and RMSE (test): val MAE 19.63 -> 17.87; test MAE 18.93 -> 17.88; test RMSE 116.65 -> 129.31 | Mixed |
| set IoU threshold for negative exemplars to 0.5 rather than extremes (ablation) | VA-Count · FSC-147 val | MAE: 0.1: 18.83; 0.9: 22.30; 0.5: 17.87 | Improved |
| set Grounding DINO logits threshold to 0.02 rather than 0.01 or 0.05 (ablation) | VA-Count · FSC-147 val | MAE: 0.01: 27.36; 0.05: 25.60; 0.02: 17.87 | Improved |
| zero-shot VA-Count versus other zero-shot methods | VA-Count · FSC-147 | MAE/RMSE: val MAE 17.87 vs 18.79 (CLIP-Count); test RMSE 129.31 vs 106.62 | Mixed |
| zero-shot VA-Count transfer without fine-tuning to crowd data | VA-Count · FSC-147 to ShanghaiTech B | MAE: 66.6 -> 40.3 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- PseCo — PseCo: Pseudo Labeling and Consistency Training for Semi-Supervised Object Detection (ECCV 2022) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- VLCounter: Text-Aware Visual Representation for Zero-Shot Object Counting (AAAI 2024 · similarity 0.90)
- T2ICount: Enhancing Cross-modal Understanding for Zero-Shot Counting (CVPR 2025 · similarity 0.89)
- A Novel Unified Architecture for Low-Shot Counting by Detection and Segmentation (NeurIPS 2024 · similarity 0.88)
- Point Segment and Count: A Generalized Framework for Object Counting (CVPR 2024 · similarity 0.88)
- DAVE - A Detect-and-Verify Paradigm for Low-Shot Counting (CVPR 2024 · similarity 0.88)
- Vision Transformer Off-the-Shelf: A Surprising Baseline for Few-Shot Class-Agnostic Counting (AAAI 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
