🎥 SportsSloMo-CVS Dataset

This repository contains the dataset presented in the paper Spatio-Temporal Difference Guided Motion Deblurring with the Complementary Vision Sensor.

The Complementary Vision Sensor (CVS), known as Tianmouc, captures synchronized RGB frames together with high-frame-rate, multi-bit spatial difference (SD, encoding structural edges) and temporal difference (TD, encoding motion cues) data within a single RGB exposure. This dataset facilitates research in RGB deblurring under extreme dynamic scenes by leveraging these complementary modalities.

📦 Dataset Overview

This is a large-scale, real-captured and pixel-level aligned dataset with diverse scenes, supporting multi-exposure and multi-form motion deblurring with CVS.

It contains:

• 98,569 training pairs
• 1,928 validation pairs
• 1,820 test pairs

Each clip folder represents one continuous motion scene.

📂 Directory Structure

Root	RGB Exposure Time (us)	Blur Length	Log	Description
Tianmouc_dataset_SportsSloMo_1160_1000_14520_1240/	14,520	11	log_1160_1000_14520_1240_filtered_fixed.csv	heavy blur
Tianmouc_dataset_SportsSloMo_1560_870_11880_1240/	11,880	9	log_1560_870_11880_1240_filtered_fixed.csv	medium-heavy blur
Tianmouc_dataset_SportsSloMo_1480_1150_9240_1240/	9,240	7	log_1480_1150_9240_1240_filtered_fixed.csv	medium blur
Tianmouc_dataset_SportsSloMo_2830_880_6600_1240/	6,600	5	log_2830_880_6600_1240_filtered_fixed.csv	light blur
Tianmouc_dataset_SportsSloMo_cop_parts/	-	-	-	GT
group_gt_parts/	-	-	-	the original images used for making the dataset

🔀 Train / Val / Test Split

Controlled by:

• val_clip_info.txt
• test_clip_info.txt

Rules:

• Train: excludes val/test clips
• Val/Test: only includes specified clips

👀 Data Preview

1. Install tianmoucv

pip install tianmoucv

2. Quick Data Preview

We provide a Jupyter Notebook to help you quickly visualize the data. (link)