日本語 
Visual Bank Co., Ltd. has launched a new Japanese Garden Image Dataset through its AI training data solution, Qlean Dataset, operated by its subsidiary Amana Images Inc. The dataset helps build and test generative AI and multimodal foundation models.
The collection features clear images of Japan’s dry landscapes and strolling gardens. It highlights special features such as ponds, stone layouts, moss, raked sand designs, and selected plants. Japanese gardens blend nature with careful design. This dataset is excellent for training AI in landscape design, spatial planning, and cultural content creation.
Each image is provided with detailed metadata, including location and subject information. This allows the dataset to be used not only for standalone visual learning, but also as part of large-scale Vision-Language Model and multimodal base model training that integrates text and imagery. The dataset is positioned for foundational model development rather than a single application use case.
Also Read: Japan Moves to Draft National AI and Robotics Strategy, Targeting Service Robot Gap
Qlean Dataset offers rights-cleared, commercially usable AI training data for both research and enterprise deployment, helping developers reduce data preparation burdens while expanding culturally specific visual understanding in AI systems.
English