Datasets › FSC147

FSC147

Introduced by Viresh Ranjan et al. in Learning To Count Everything16 Apr 2021 archive 2025-07-28

We introduce a dataset of 147 object categories containing over 6000 images that are suitable for the few-shot counting task. We collected and annotated images ourselves. Our dataset consists of 6135 images across a di- verse set of 147 object categories, from kitchen utensils and office stationery to vehicles and animals. The object count in our dataset varies widely, from 7 to 3731 objects, with an average count of 56 objects per image. In each image, each object instance is annotated with a dot at its approxi- mate center. In addition, three object instances are selected randomly as exemplar instances; these exemplars are also annotated with axis-aligned bounding boxes.

Benchmarks archive 2025-07-28

All 4 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

Papers archive 2025-07-28

22 shown of 22 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 58. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
SAVE: Self-Attention on Visual Embedding for Zero-Shot Generic Object Counting 1 1 10 Feb 2025 not harvested
A Novel Unified Architecture for Low-Shot Counting by Detection and Segmentation 1 3 27 Sep 2024 not harvested
GCA-SUNet: A Gated Context-Aware Swin-UNet for Exemplar-Free Counting 0 2 18 Sep 2024 not harvested
CountGD: Multi-Modal Open-World Counting 2 1 5 Jul 2024 ran 4 of 9 samples (5 unverified)
Learning Spatial Similarity Distribution for Few-shot Object Counting 1 1 20 May 2024 ran 4 of 10 samples (6 unverified; 10 pointer-only for licence)
DAVE -- A Detect-and-Verify Paradigm for Low-Shot Counting 1 3 25 Apr 2024 ran 16 of 18 samples (2 unverified)
OmniCount: Multi-label Object Counting with Semantic-Geometric Priors 0 2 8 Mar 2024 not harvested
Point Segment and Count: A Generalized Framework for Object Counting 1 1 1 Jan 2024 not harvested
Semantic Generative Augmentations for Few-Shot Counting 1 2 26 Oct 2023 not harvested
Training-free Object Counting with Prompts 1 1 30 Jun 2023 not harvested
Open-world Text-specified Object Counting 1 1 2 Jun 2023 not harvested
Vision Transformer Off-the-Shelf: A Surprising Baseline for Few-Shot Class-Agnostic Counting 1 1 8 May 2023 not harvested
Scale-Prior Deformable Convolution for Exemplar-Guided Class-Agnostic Counting 1 1 21 Dec 2022 not harvested
A Low-Shot Object Counting Network With Iterative Prototype Adaptation 1 2 15 Nov 2022 ran 0 of 2 samples (2 unverified)
CounTR: Transformer-based Generalised Visual Counting 1 2 29 Aug 2022 ran 1 of 4 samples (3 unverified)
Few-shot Object Counting and Detection 1 2 22 Jul 2022 ran 0 of 4 samples (4 unverified)
Learning to Count Anything: Reference-less Class-agnostic Counting with Weak Supervision 2 2 20 May 2022 ran 1 of 5 samples (4 unverified)
Represent, Compare, and Learn: A Similarity-Aware Framework for Class-Agnostic Counting 1 1 16 Mar 2022 ran 0 of 1 samples (1 unverified)
Few-shot Object Counting with Similarity-Aware Feature Enhancement 1 1 22 Jan 2022 not harvested
Exemplar-free Class Incremental Learning via Discriminative and Comparable One-class Classifiers 1 1 5 Jan 2022 not harvested
Object Counting: You Only Need to Look at One 0 1 11 Dec 2021 not harvested
Learning To Count Everything 1 2 16 Apr 2021 ran 0 of 1 samples (1 unverified)

Dataset loaders archive 2025-07-28

1 loader as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

MIT LICENSE

Modalities archive 2025-07-28

Languages archive 2025-07-28

Variants archive 2025-07-28

  • FSC147

1 variant name, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections