Turing Announces Two Papers Accepted at CVPR 2026, the World’s Premier Computer Vision Conference
Turing Inc. (Headquarters: Ota-ku, Tokyo; CEO: Issei Yamamoto; hereinafter “Turing”) is pleased to announce that two papers authored by its Research and Development team have been accepted for presentation at the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2026.
CVPR is widely recognized as one of the world’s most prestigious international conferences in the field of computer vision. Each year, numerous papers are submitted by leading universities, research institutions, and corporations globally, with final selections made through a rigorous peer-review process. Covering a broad spectrum of topics—including image understanding and multimodal learning—CVPR is a flagship academic event that defines the cutting edge of the industry.
Accepted Paper #1
Text-Printed Image: Bridging the Image-Text Modality Gap for Text-centric Training of Large Vision-Language Models
Overview
This paper proposes “Text-Printed Image,” a novel method for enhancing Large Vision-Language Models (LVLMs) using only text descriptions, without the need for real-world images. The core of this approach lies in rendering input text directly onto a blank canvas to create a synthetic image. This allows the model to integrate semantic text information into existing image-learning pipelines while preserving the original meaning of the text. Through extensive evaluation across multiple models and benchmarks, we confirmed that this method enables more effective text-centric learning compared to using images generated by diffusion models.
Paper Link (arXiv): https://arxiv.org/abs/2512.03463
Tech Blog: https://zenn.dev/turing_motors/articles/017e6d6253632f
Accepted Paper #2
P2GS: Physical Prior-guided Gaussian Splatting for Photometrically Consistent Urban Reconstruction
Overview
This research introduces “P2GS,” a novel method for reconstructing high-quality 3D spaces from images captured under varying conditions. Conventional approaches often struggle with inconsistencies in brightness and color across cameras, resulting in unnatural lighting contrasts and color shifts in reconstructed 3D models. P2GS addresses these challenges by more accurately estimating the underlying illumination of real-world environments. This enables consistent 3D reconstruction across different cameras and lighting conditions, providing a stable and realistic virtual foundation for applications such as autonomous driving simulation.
Impact and Future Outlook
The two papers accepted this year contribute to streamlining the development process through advancements in large-scale model training and high-fidelity simulation environments. These research achievements play a critical role in building safer and more reliable autonomous driving systems. Turing remains committed to cutting-edge R&D and will continue strengthening the technological foundation necessary to realize fully autonomous driving.
About Turing
Turing is a deep-tech startup dedicated to the development of fully autonomous driving. We are simultaneously developing End-to-End (E2E) autonomous driving AI—which integrates environmental perception, path planning, and vehicle control into a single system—alongside Large Foundation Models capable of understanding human social norms, backgrounds, and contexts. By integrating these core technologies, Turing aims to achieve “fully autonomous driving,” where vehicles can operate on behalf of humans under all conditions.
Company Overview
Company Name: Turing Inc.（Japanese：チューリング）
Headquarters: East Tower 4F, Gate City Ohsaki, 1-11-2 Ohsaki, Shinagawa-ku, Tokyo
CEO: Issei Yamamoto
Founded: August 2021
Business: Development of fully autonomous driving technologies
Website: https://tur.ing/
Careers
Turing is hiring individuals passionate about transforming the world through fully autonomous Japanese driving technologies. Please visit our careers page to learn more. We also host regular events such as open office sessions and Tech Talks.
Media Contact
PR Representative (Abe)：pr@turing-motors.com