EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

ABACUS: Adapting Unified Foundation Model for Count Image Understanding and Generation

Anindya Mondal, Sauradip Nag, Anjan Dutta

ACM Transactions on Graphics · 09.11.2026 · 45 (6) · 1-14

DOI: 10.1145/3842572 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Graphics, online erschienen am 09.11.2026. Verfasst von Anindya Mondal und 2 weiteren Autor:innen. Die Arbeit nennt 106 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We present ABACUS, a unified vision-language model that jointly addresses object counting, crowd counting, referring-expression counting, and count-faithful image generation within a single 3B-parameter model. ABACUS introduces three contributions: density-aware adaptive zooming paired with an objectness map from multi-head self-attention decomposition to spatially ground count predictions; a boundary-aware count policy trained via GRPO with nested local, boundary, and global rewards to eliminate over- and undercounting at crop boundaries; and a cycle-consistent GRPO strategy in which the frozen understanding branch scores generated candidates on count-deviation and aesthetic quality, closing the understanding-generation synergy gap without any external critic or annotation. ABACUS achieves state-of-the-art results across seven benchmarks spanning object counting (FSC-147, CARPK), crowd counting (ShanghaiTech A/B), referring-expression counting (REC-8K), count-faithful generation (CoCo-Count, T2I-CompBench, GenEval), and count reasoning (CountQA), surpassing both task-specific specialists and larger generalist models. Project page is at https://mondalanindya.github.io/pages/ABACUS/.

Zitieren

APA

Mondal, A., Nag, S., & Dutta, A. (2026). ABACUS: Adapting Unified Foundation Model for Count Image Understanding and Generation. ACM Transactions on Graphics, 45(6), 1-14. https://doi.org/10.1145/3842572

Vancouver

Mondal A, Nag S, Dutta A. ABACUS: Adapting Unified Foundation Model for Count Image Understanding and Generation. ACM Transactions on Graphics. 2026;45(6):1-14. doi:10.1145/3842572

BibTeX
@article{Mondal2026,
  author = {Mondal, Anindya and Nag, Sauradip and Dutta, Anjan},
  title = {ABACUS: Adapting Unified Foundation Model for Count Image Understanding and Generation},
  journal = {ACM Transactions on Graphics},
  year = {2026},
  volume = {45},
  number = {6},
  pages = {1--14},
  doi = {10.1145/3842572}
}
Kostenlos starten →