Qlean Dataset Launches "Japanese Daily Household Activity Video Dataset"
Visual Bank Inc., through its subsidiary Amana Images, has launched the "Japanese Daily Household Activity Video Dataset" via its AI learning data solution "Qlean Dataset." This dataset is optimized for improving visual recognition algorithms in home robotics and learning advanced action recognition models through video analysis, featuring diverse household tasks recorded in realistic Japanese living environments.
📋 Article Processing Timeline
- 📰 Published: April 1, 2026 at 23:00
- 🔍 Collected: April 1, 2026 at 16:47
- 🤖 AI Analyzed: April 21, 2026 at 23:39 (486h 51m after Collected)
Visual Bank Inc. (Minato-ku, Tokyo; Representative Director and CEO: Masayuki Nagai), through its subsidiary Amana Images Co., Ltd., has launched the "Japanese Daily Household Activity Video Dataset" as part of its AI learning data solution "Qlean Dataset." This dataset is optimized for improving visual recognition algorithms in home robotics and learning advanced action recognition models (Action Recognition) through video analysis.
This dataset consists of video data and metadata recording the daily household chores of Japanese people in real Japanese living environments. It covers not just isolated actions but a wide range of tasks, from intricate hand operations like chopping and plating food to full-body movements such as vacuuming, hanging, and folding laundry. By including diverse clothing variations—from young to elderly, with or without aprons, and different sleeve lengths—it enables the construction of robust models that are unaffected by visual noise and changes in the real world.
The filming locations are limited to real living spaces within homes, such as kitchens, living rooms, and laundry areas. This allows for deep learning of "site-specific visual information" that is difficult to reproduce in simulated environments, such as natural light penetration and reflections from indoor lighting. In addition to mid-shots that provide an overview of the entire action, close-ups focusing on hand movements are recorded from multiple angles. This serves as practical training data for analyzing object manipulation involving physical contact (V-O Interaction) and for research and development of advanced context-aware action prediction.
This data is offered as one of the original data lineups for AI development, "AI Data Recipe," under the Qlean Dataset. It strongly supports AI development aimed at social implementation, from the deployment of next-generation domestic service robots to the digitalization of lifestyle habits through video analysis. Visual Bank and Amana Images will continue to support AI research and development that accurately understands and analyzes Japanese living environments by providing dynamic structural data capturing Japanese living spaces.
## Overview of the "Japanese Daily Household Activity Video Dataset" Now Available
- Data Type: Video
- Subject Attributes:
- Actions: Cooking (chopping/plating ingredients), cleaning, laundry (hanging/folding), water-related tasks (dishwashing/handwashing), others (flower arranging, tidying up, etc.)
- Attributes: Men and women of all ages, various clothing (wearing aprons, diverse clothing variations with different sleeve lengths like long sleeves/short sleeves)
This dataset consists of video data and metadata recording the daily household chores of Japanese people in real Japanese living environments. It covers not just isolated actions but a wide range of tasks, from intricate hand operations like chopping and plating food to full-body movements such as vacuuming, hanging, and folding laundry. By including diverse clothing variations—from young to elderly, with or without aprons, and different sleeve lengths—it enables the construction of robust models that are unaffected by visual noise and changes in the real world.
The filming locations are limited to real living spaces within homes, such as kitchens, living rooms, and laundry areas. This allows for deep learning of "site-specific visual information" that is difficult to reproduce in simulated environments, such as natural light penetration and reflections from indoor lighting. In addition to mid-shots that provide an overview of the entire action, close-ups focusing on hand movements are recorded from multiple angles. This serves as practical training data for analyzing object manipulation involving physical contact (V-O Interaction) and for research and development of advanced context-aware action prediction.
This data is offered as one of the original data lineups for AI development, "AI Data Recipe," under the Qlean Dataset. It strongly supports AI development aimed at social implementation, from the deployment of next-generation domestic service robots to the digitalization of lifestyle habits through video analysis. Visual Bank and Amana Images will continue to support AI research and development that accurately understands and analyzes Japanese living environments by providing dynamic structural data capturing Japanese living spaces.
## Overview of the "Japanese Daily Household Activity Video Dataset" Now Available
- Data Type: Video
- Subject Attributes:
- Actions: Cooking (chopping/plating ingredients), cleaning, laundry (hanging/folding), water-related tasks (dishwashing/handwashing), others (flower arranging, tidying up, etc.)
- Attributes: Men and women of all ages, various clothing (wearing aprons, diverse clothing variations with different sleeve lengths like long sleeves/short sleeves)