SUPPLEMENTARY INFORMATION:

Background: In 2018, thirty Federal organizations came together under the auspices of the NITRD VIA Team to develop an action plan for providing direction, coherence, and consensus across federal R&D efforts.

Since publication of the VIA R&D Action Plan in 2020, government agencies have invested in computer vision research from foundational efforts to the development of systems that support agency missions and programs. As a reference, NITRD supplements to the President's annual budget requests collect information on government spending in artificial intelligence (AI), part of which includes computer vision research, showing close to $800 million in FY 2020, close to $1 billion in FY 2021, and over $1 billion in requests for FY 2022, as noted in Artificial Intelligence R&D Investments FYs 2018-2022.

At the same time, private sector funding on R&D of applications for video and image analysis continues to dwarf government funding in almost every domain. Examples range from combining computer vision with robotics for increased automation in agriculture to increasing the use of biometrics to provide secure access to consumer devices. Government can take advantage of the results of private sector R&D through an open marketplace. Still, the need remains for Federal research investment, given that there is a strong public benefit but little to no opportunity for building a private sector business case.

Along with the advances in research, development, and technology implementation over the past several years, there have also been changes in societal awareness of, expectations for, and the use of computer vision. For example, at the Federal level, Executive Order 13960, Promoting the Use of Trustworthy Artificial Intelligence in the Federal Government, reflects the importance of human interaction with AI technologies, including video and image analytics.

The VIA R&D Action Plan lists the following six strategic goals and objectives:

Invest in foundational research that applies to multiple domains

Develop new, agile, and effective R&D methodologies

Improve communication and coordination across government agencies

Improve outreach and knowledge sharing between the public and private sectors

Develop and sustain an effective and diverse technical workforce and a robust research community

Develop and promulgate standards and best practices that support integration of R&D

Objectives: NITRD NCO seeks input on potential revisions to the VIA R&D Action Plan to reflect changes in technology and the socio-technical environment over the past four years. Responders are asked to answer the following questions in response to the RFI:

1. What R&D has taken place relevant to each goal since the VIA R&D Action Plan was published in March 2020? How does R&D reduce or change the need for the Federal Government to continue to pursue each goal?

2. What societal changes have taken place that would impact the need for the Federal Government to continue to pursue each goal?

3. With R&D advances and changes in the socio-technical environment, what additional goal(s) should the Federal Government consider?

References

Federal Video and Image Analytics Research and Development Action Plan (March 2020): https://www.nitrd.gov/​pubs/​RD-Opportunities-in-Video-Image-Analytics-2020.pdf.

NITRD NAIIO Supplement to the President's FY2022 Budget: https://www.nitrd.gov/​pubs/​FY2022-NITRD-NAIIO-Supplement.pdf.

Executive Order 13960, Promoting the Use of Trustworthy Artificial Intelligence in the Federal Government (December 2020): https://www.federalregister.gov/​d/​2020-27065.

87 FR 42212