G-250: Color Triple Intersection Red Data Generator

Generates synthetic tasks featuring Venn diagrams with three circles. The task is to identify the triple intersection region (where all three circles overlap) and color it red.

Each sample pairs a task (first frame + prompt describing what needs to happen) with its ground truth solution (final frame showing the result + video demonstrating how to achieve it). This structure enables both model evaluation and training.

---

📌 Basic Information

Property	Value
Task ID	G-250
Task	Color Triple Intersection Red
Category	Perception
Resolution	1024×1024 px
FPS	16 fps
Duration	~3.75 seconds
Output	PNG images + MP4 video

---

🚀 Usage

Installation

# Clone the repository
git clone https://github.com/Jiaqi-Gong/Gong_VBVR_Data.git
cd Gong_VBVR_Data/G-250_color_triple_intersection_red_data-generator

# Install dependencies
pip install -r requirements.txt

Generate Data

# Generate 100 samples
python examples/generate.py --num-samples 100

# Generate with specific seed
python examples/generate.py --num-samples 100 --seed 42

# Generate without videos
python examples/generate.py --num-samples 100 --no-videos

# Custom output directory
python examples/generate.py --num-samples 100 --output data/my_output

Command-Line Options

Argument	Type	Description	Default
--num-samples	int	Number of samples to generate	100
--seed	int	Random seed for reproducibility	Random
--output	str	Output directory	data/questions
--no-videos	flag	Skip video generation	False

---

📖 Task Example

Prompt

A Venn diagram of circles is shown. Identify the region that lies in all three of the first three circles (triple intersection) and color that region red. Do not change anything else.

Visual

Initial Frame Three overlapping circles with outlines	Animation Triple intersection region fades to red	Final Frame Triple intersection colored red

---

📖 Task Description

Objective

Identify the region where all three circles in a Venn diagram overlap simultaneously (triple intersection) and fill it with red color. This generator involves color modification tasks, adding red color to a specific region while preserving all other visual properties.

Task Setup

• Circle Count: Exactly 3 circles (prevents ambiguity of multiple triple intersections)
• Circle Appearance: Outlined circles with no fill (outlines in various colors)
• Circle Size: Radius 22-28% of image size (225-287 pixels for 1024×1024)
• Outline Colors: 40 distinct colors (excluding red) for circle outlines
• Outline Width: 2-6 pixels (randomly selected for variation)
• Margin: Circles stay at least 50px from image edges
• Answer Region: Triple intersection filled with red (RGB: 255, 0, 0)

Key Features

• Exact triple overlap: All three circles must overlap in a common region
• No fill in circles: Only outlines are visible initially (makes intersection clear)
• Varied positions: Circles placed randomly while ensuring triple overlap exists
• Smooth animation: Red color fades in over the triple intersection (10 hold + 40 transition + 10 hold frames)
• Unambiguous answer: Only one triple intersection region per image
• Scalable design: 40 outline colors × 5 outline widths × many positions = high uniqueness

---

📦 Data Format

data/questions/color_triple_intersection_red_task/color_triple_intersection_red_00000000/
├── first_frame.png      # Initial state (outlined circles)
├── final_frame.png      # Goal state (triple intersection red)
├── prompt.txt           # Task instructions
├── ground_truth.mp4     # Solution video (16 fps)
└── question_metadata.json # Task metadata

File specifications: Images are 1024×1024 PNG. Videos are MP4 at 16 fps, approximately 3.75 seconds long (60 frames total).

---

🏷️ Tags

venn-diagram set-theory intersection spatial-reasoning region-identification

---