#Japan #Tokyo #Hugging Face #Qlean Dataset #Visual Bank

Visual Bank's Initiative to Provide Free Data to Academia

In a significant move for the academic community, Visual Bank Inc. has announced that it will offer its Qlean Dataset, a specialized resource for AI training, free of charge through Hugging Face Hub. This initiative is aimed at supporting researchers affiliated with universities and research institutions, enabling them to access rights-cleared datasets that can greatly enhance AI research efforts. Under a specific Academic Research License, participating researchers will not only have the ability to view and evaluate the data but will also be able to train, fine-tune, and assess AI/ML models for their academic projects, all at no cost.

Expanding the Academia Support Program

Visual Bank's decision to expand its Academia Support Program, launched in 2025, builds upon its proven track record of supporting more than ten research laboratories. The newly accessible datasets will empower AI researchers and developers from around the globe to utilize Japanese-specific data, which is often underrepresented in publicly available datasets. This initiative is particularly timely, as the field of AI research becomes increasingly competitive, and the availability of high-quality training data is imperative for achieving successful outcomes.

Addressing Data Scarcity in Japan

The need for diverse, high-quality datasets has intensified as AI research evolves. Japan has been facing a structural deficit of native data, leading to a concentration of research efforts on datasets originating from other countries. Specifically, statistics show that of the datasets available on Hugging Face Hub, only about 3,000 are Japanese, representing approximately 0.3% of the total, compared to the 85,000 English datasets available. Furthermore, most of the Japanese data is derived from multilingual datasets created outside of Japan, constraining researchers' ability to differentiate their work based on unique cultural and linguistic contexts.

Visual Bank’s effort aims to address this issue by offering data that specifically reflects the diversity of Japanese culture, language, and societal nuances. Among the resources included in the initial dataset lineup are voice conversation datasets featuring emotional nuances, separated audio channels for private dialogue, diverse facial images representative of various demographics, and recruitment self-promotion videos created by students.

Encouraging a Collaborative Research Environment

Moreover, researchers using the Qlean Dataset are permitted to publish their findings, provided they are conducted for non-commercial and academic purposes. This aspect fosters a collaborative research environment, enabling researchers to contribute to the academic community while driving innovation in AI. By allowing the trained models to be shared within the confines of academic work, researchers will be further encouraged to explore and exploit the potential of the datasets they now have access to.

The prominent role of the industry in AI research has also been increasing, which was highlighted in a Stanford report that indicated that about 90% of major AI models released in 2024 were developed by industry leaders—a jump from 60% in 2023. This trend underscores the need for academic researchers to leverage unique datasets to innovate distinct and valuable contributions to the field of AI.

Future Directions and Continued Support

Visual Bank's collaboration with Hugging Face Hub marks a turning point in access to valuable datasets within the academic landscape. Moving forward, the company plans to continually add a variety of datasets—including audio, image, and video formats—to further enhance the research capabilities of academia. Additionally, Visual Bank remains committed to providing personalized support through its existing Academia Support Program, adapting its offerings to meet the unique needs of individual researchers and their respective studies.

This groundwork sets the stage for a flourishing research environment, enabling scholars to incorporate Japan-specific data into their work and fostering the advancement of AI research that resonates specifically with Japanese culture.

Conclusion

As this initiative unfolds, Visual Bank demonstrates a proactive stance toward nurturing the future of AI discourse in academia. By enabling unlimited access to rights-cleared data, they are not only empowering the next generation of researchers but also contributing significantly to the global dialogue surrounding AI development and its ethical implications. Researchers, educators, and students alike can anticipate transformative advancements in the field as a result of this pivotal step toward democratizing access to essential datasets.

To explore the Qlean Dataset, visit the Hugging Face Hub at Hugging Face Hub page and discover the myriad resources available for academic research.