The program committee of the 2026 Conference on Computer Vision and Pattern Recognition (CVPR) has released the details of this year's technical program, reporting a record 16,092 paper submissions. The figure represents a 24 percent increase over 2025 and underscores the breakneck pace of research in computer vision, embodied intelligence and multimodal AI.
A record year for submissions
Co-sponsored by the IEEE Computer Society and the Computer Vision Foundation, CVPR remains one of the most selective venues in the field. Through a rigorous peer-review process, roughly one-quarter of the submissions were accepted, resulting in 4,089 paper presentations on the program.
"CVPR submissions have more than doubled over the past five years, but the acceptance rate has remained highly competitive, consistently in the low-to-mid 20 percent range," said Alexander G. Schwing, associate professor of electrical and computer engineering at the University of Illinois Urbana-Champaign and a CVPR 2026 program co-chair. "While AI demand has fueled expansive research, CVPR remains one of the most selective and prestigious technical events in the field."
Where the research is concentrated
The committee received the highest number of submissions in image and video synthesis and generation; vision, language and reasoning; multi-modal learning; 3-D from multiview and sensors; and medical and biological vision and cell microscopy. The mix mirrors broader momentum in agentic and multimodal AI systems and the surge of capital flowing into physical AI and robotics startups.
"As fundamental concepts of computer vision permeate new applications, we're seeing a rise in submitted research that corresponds with particular disciplines," said Chen Change Loy, president's chair professor at Nanyang Technological University, Singapore, and a CVPR 2026 program co-chair. He noted that emphasis on medical and biological vision grew substantially this year, though the work remains in nascent stages.
Award-candidate papers
Among the accepted papers flagged as award candidates is NitroGen, a vision-action foundation model for generalist gaming agents from a team including Nvidia, Stanford University, the California Institute of Technology, the University of Chicago and the University of Texas at Austin, trained on 40,000 hours of gameplay videos across more than 1,000 games. Other highlighted work includes a diffusion framework for photorealistic mobile bokeh rendering, a black-box membership inference attack framework for fine-tuned diffusion models from the University of Virginia, and R2Seg, a training-free tumor-segmentation method developed by researchers at Carnegie Mellon University, the University of Cambridge, Zhejiang University, ETH Zurich and the University of Illinois Urbana-Champaign.
Why it matters
CVPR's proceedings consistently rank among the most cited in AI, taking the number two spot in Google's 2025 Scholar Metrics. The record submission count signals that demand for advances in perception, generation and reasoning continues to accelerate, even as breakthroughs in adjacent fields such as AI-driven mathematical research capture headlines. The full technical program will be presented at the conference, with abstracts available to registered attendees.
Reporting based on coverage from Robotics & Automation News and the CVPR 2026 program committee (IEEE Computer Society and the Computer Vision Foundation).