README.md

---
language:
- en
license: apache-2.0
task_categories:
- visual-question-answering
- image-classification
- image-to-text
pretty_name: FineSightBench
size_categories:
- 1K<n<10K
tags:
- VLM-evaluation
- fine-grained-visual-perception
- fine-grained-visual-reasoning
- text-in-the-wild
- scene-text-recognition
splits:
- name: perception
  num_examples: 4200
- name: reasoning
  num_examples: 3920
dataset_info:
  features:
  - name: image
    dtype: image
  - name: image_id
    dtype: string
  - name: task_type
    dtype: string
  - name: question
    dtype: string
  - name: answer
    dtype: string
  - name: difficulty
    dtype: string
  - name: metadata
    dtype: string
  splits:
  - name: perception
    num_bytes: 227180504
    num_examples: 4200
  - name: reasoning
    num_bytes: 487360638
    num_examples: 3920
  download_size: 708023982
  dataset_size: 714541142
configs:
- config_name: default
  data_files:
  - split: perception
    path: data/perception-*
  - split: reasoning
    path: data/reasoning-*
---

# FineSightBench

**FineSightBench** is a fine-grained visual benchmark for evaluating Vision-Language Models (VLMs) on pixel-level perception and reasoning tasks. It combines two complementary image regimes:

1. **Synthetic canvas** — controlled white-background images with precisely-sized geometric/semantic targets (letters, animals, shapes, blocks, dots).
2. **Text in the wild** (SynthText-style) — English words rendered onto real natural-scene photographs from the [SynthText](https://github.com/ankush-me/SynthText) `bg_img` set, with **pixel-accurate control of character cap-height**.

All images are **448 × 448 px**. The primary difficulty axis is the **target pixel size** (cap-height for text), swept over `[4, 8, 12, 16, 24, 32, 48]`, mapped to `extreme / hard / medium / easy`.

## Dataset Summary

| Split | #Samples | #Task types | Regimes |
|-------|---------:|:-----------:|---------|
| `perception` | 4 200 | 6 | synthetic canvas + text-in-the-wild |
| `reasoning`  | 3 920 | 6 | synthetic canvas + text-in-the-wild |

## Dataset Structure

### `perception` split — 4 200 samples

Single-target identification tasks. 700 samples per task, 100 samples per pixel size × 7 sizes.

| `task_type` | Description | Source |
|-------------|-------------|--------|
| `letter_recognition` | Identify a rendered uppercase letter (A–Z) | synthetic canvas |
| `animal_recognition` | Identify an animal silhouette (cat/dog/fish/bird/rabbit/turtle) | synthetic canvas |
| `shape_recognition` | Identify a geometric shape (circle/triangle/square/star/diamond/pentagon/hexagon/cross) | synthetic canvas |
| `block_recognition` | Detect / count square blocks | synthetic canvas |
| `color_block_recognition` | Identify the color of a block | synthetic canvas |
| `text_recognition` | Read a single English word overlaid on a natural scene | **text in the wild** |

### `reasoning` split — 3 920 samples

Chain-reasoning tasks requiring counting, ordering, and spatial reasoning across multiple targets.

| `task_type` | Description | Source |
|-------------|-------------|--------|
| `spatial_chain` | List all objects left→right or top→bottom | synthetic canvas |
| `comparison_chain` | List all objects smallest→largest by size | synthetic canvas |
| `counting_chain` | Count objects per type + total | synthetic canvas |
| `blur_chain` | Count objects on a blurred/textured background | synthetic canvas |
| `text_reading_chain` | Read multiple overlaid words in left→right / top→bottom order | **text in the wild** |
| `text_counting_chain` | Total word count + # words containing a queried letter | **text in the wild** |

### Difficulty levels

| Difficulty | Target / cap-height |
|------------|---------------------|
| `extreme`  | ≤ 5 px |
| `hard`     | 6–12 px |
| `medium`   | 13–24 px |
| `easy`     | 25–48 px |

## Fields

| Field | Type | Description |
|-------|------|-------------|
| `image`      | Image    | 448×448 PNG |
| `image_id`   | string   | Unique identifier (encodes task, size, count) |
| `task_type`  | string   | See tables above |
| `question`   | string   | Prompt for the VLM (asks for a structured JSON answer) |
| `answer`     | string   | Ground-truth answer. JSON-encoded (see below) |
| `difficulty` | string   | `easy` / `medium` / `hard` / `extreme` |
| `metadata`   | string   | JSON with canvas size, target pixel size, positions, colors, bounding boxes, sub-answers, etc. |

### Answer schemas (examples)

| Task | Answer JSON |
|------|-------------|
| `letter_recognition` | `{"letter": "H"}` |
| `animal_recognition` | `{"animal": "rabbit"}` |
| `shape_recognition`  | `{"shape": "triangle"}` |
| `color_block_recognition` | `{"color": "blue"}` |
| `text_recognition`   | `{"word": "HOME"}` |
| `spatial_chain`      | `{"objects": ["red A", "blue K", ...]}` |
| `comparison_chain`   | `{"objects": ["blue dog", "magenta bird"]}` |
| `counting_chain`     | `{"counts": {"red": 2, "blue": 1}, "total": 3}` |
| `blur_chain`         | `{"counts": {"circle": 1, "square": 2}, "total": 3}` |
| `text_reading_chain` | `{"words": ["HOME", "CITY", "EXIT"]}` |
| `text_counting_chain`| `{"total": 6, "with_letter": 3}` |

## Usage

```python
from datasets import load_dataset

ds = load_dataset("Volavion/FineSightBench")
print(ds)
# DatasetDict({
#     perception: Dataset({features: [...], num_rows: 4200}),
#     reasoning:  Dataset({features: [...], num_rows: 3920})
# })

sample = ds["perception"][0]
sample["image"].show()
print(sample["question"])
print(sample["answer"])      # JSON string, e.g. '{"letter": "A"}'
```

Filter by task or difficulty:

```python
text_subset = ds["perception"].filter(lambda x: x["task_type"] == "text_recognition")
extreme    = ds["perception"].filter(lambda x: x["difficulty"] == "extreme")
```

## Design Philosophy

* **Pixel-size is the primary difficulty axis.** Targets (objects or characters) are rendered at exact cap-heights across `[4, 8, 12, 16, 24, 32, 48]` px so that the same semantic task can be probed from *easily readable* to *near-imperceptible* scales on a single fixed 448×448 canvas.
* **Controlled composition.** Every sample exposes pixel-precise target positions, bounding boxes, colors (with RGB), and sub-answers in `metadata`, enabling per-task, per-size, per-color, and positional analyses.
* **Two image regimes.** The synthetic canvas removes distribution confounders, while the SynthText-style text-in-the-wild regime stresses models with the same text task on varied, real photographs.

## Generation

Generated with the [FineSightBench repository](https://github.com/Volavion/FineSightBench):

```bash
python generate.py perception --output data/full_perception
python generate.py reasoning  --output data/full_reasoning
python generate.py textwild   --split all --output data   # merges text-in-the-wild tasks
```

**Text-in-the-wild backgrounds**: the first ~1 500 JPEGs from the SynthText `bg_img.tar.gz` set ([mirror](https://thor.robots.ox.ac.uk/scenetext/preproc/bg_img.tar.gz)) are center-cropped and resized to 448×448. Text glyphs use system sans-serif fonts; cap-height is calibrated per render to match the requested pixel size exactly.

## Citation

If you use FineSightBench, please cite the repository and the SynthText background source:

```bibtex
@misc{finesightbench2026,
  title  = {FineSightBench: Fine-grained Visual Perception \& Reasoning Benchmark for VLMs},
  year   = {2026},
  url    = {https://huggingface.co/datasets/Volavion/FineSightBench}
}

@inproceedings{Gupta16,
  author    = {A. Gupta and A. Vedaldi and A. Zisserman},
  title     = {Synthetic Data for Text Localisation in Natural Images},
  booktitle = {IEEE Conference on Computer Vision and Pattern Recognition},
  year      = {2016}
}
```

## License

Apache-2.0 for the FineSightBench benchmark code, annotations, and synthetic images. The natural-scene backgrounds for the text-in-the-wild tasks are derived from the SynthText `bg_img` set; please refer to the original SynthText dataset for the background-image license/terms.