4/1/2026
New Video Dataset for Japanese Human Activity Recognition (HAR)
Visual Bank Inc. (Minato-ku, Tokyo; Saneyuki Nagai, Representative Director and CEO), through its subsidiary amana images inc., is pleased to announce the release of the "Japanese Daily Household Activity Video Dataset" as part of its AI training data solution, "Qlean Dataset." This dataset is specifically optimized for improving visual recognition algorithms in home robotics and training advanced Action Recognition models through video analysis.
The dataset consists of video footage and rich metadata capturing Japanese subjects performing routine household chores within authentic domestic environments in Japan. Rather than simple action snippets, the data covers end-to-end processes: from meticulous manual tasks like chopping and plating ingredients to full-body movements such as vacuuming and folding laundry. By including a diverse demographic—from young adults to the elderly—and varying attire (e.g., presence of aprons, different sleeve lengths), the dataset enables the construction of robust models capable of handling the visual noise and variability found in the real world.
Filming was conducted exclusively in actual residential spaces, including kitchens, living rooms, and laundry areas. This allows for deep learning of field-specific visual information—such as natural light incursion and indoor lighting reflections—that is difficult to replicate in simulated environments. The footage features multi-angle coverage, ranging from middle shots capturing overall movement to close-ups focused on manual dexterity. This serves as high-quality ground truth data for analyzing Video-Object Interaction (VOI) and developing research for context-aware activity prediction.
This release is part of "AI Data Recipe," the proprietary lineup of AI development data provided by Qlean Dataset. It is designed to empower AI development at the social implementation stage, from deploying next-generation domestic service robots to digitizing lifestyle habits through video analytics. Visual Bank and amana images remain committed to supporting the research and development of AI that accurately understands and analyzes Japanese living environments by providing dynamic, structured data of physical assets.
Dataset Overview: Japanese Daily Household Activity Video Dataset
Data Type: | Video |
|---|---|
Subject Attributes: | ・Actions: Cooking (chopping, plating), cleaning, laundry (hanging, folding), wet areas (dishwashing, handwashing), and others (arranging flowers, tidying up, etc.). |
Format: | mp4 / mov |
Recording Environment: | * Interiors of typical Japanese homes (kitchen, living room, laundry space, window-side, etc.).Shot under natural and indoor lighting. |
Angles: | Multi-angle configuration, including middle shots for full action and close-ups for manual work. |
Metadata: | Included |
Sample Details: |
Potential Use Cases
【Research & Academia】
Validation of Activity Prediction Models in Mixed First-Person/Third-Person Perspectives
By using footage with complex backgrounds and occlusions typical of real homes, researchers can validate the accuracy of algorithms predicting the next likely action of a person or conduct object recognition studies using semantic segmentation.
【Industrial & Corporate】
Development of Task Execution Algorithms for Autonomous Home Robots
Analysis of household movement patterns (e.g., cleaning or tidying) enables the development of path planning for robots to navigate without interfering with humans. It also facilitates training models for robotic manipulators to perform flexible tasks like folding laundry or washing dishes.
【Social Implementation & Care】
Optimization of Monitoring Systems for the Elderly and Caregivers
By incorporating non-routine actions like "arranging flowers" or "organizing," anomaly detection AI can be trained to accurately distinguish "normal daily activities" from emergencies such as falls or sudden health changes, providing a reliable benchmark for implementation testing.
About Qlean Dataset
Qlean Dataset is a commercially cleared AI training data solution provided by amana images Inc., a subsidiary of Visual Bank Group. The platform offers diverse data formats including image, video, audio, 3D, and text, as well as a specialized AI Data Recipe lineup developed through collaborations with major media organizations and data rights holders.
URL:https://qleandataset.visual-bank.co.jp/en
About Visual Bank Inc.
Visual Bank Group is a technology company developing data infrastructure and AI solutions that support advanced AI development. The company operates THE PEN, an AI tool for manga creators, and its subsidiary, amana images Inc., provides commercial digital content and AI training data solutions, including Qlean Dataset. Visual Bank is also a selected participant in GENIAC, a Japanese government initiative supporting the advancement of next generation AI technologies.
CEO: Saneyuki Nagai
Website:https://visual-bank.co.jp/en
