Recently, the prestigious CVPR (Computer Vision and Pattern Recognition Conference), the world’s leading event in the field of computer vision, was held in Seattle, USA. This year, OPPO made a significant contribution by presenting several of its latest research papers that were selected for presentation, covering a wide range of AI-related topics.
In addition, OPPO’s flagship AI phone, the Find X7 Ultra, and the recently launched OPPO Reno12, which showcases numerous innovative AI phone features, were showcased at the conference. With this, OPPO continues to influence the global development and democratization of AI phones.
“OPPO has been actively participating in CVPR every year, and this year’s achievements are particularly notable,” said Billy Zhang, President of OPPO’s overseas marketing, sales and service department.
According to him, artificial intelligence has always been a key area of investment for the company, and the growing trend for AI phones allows us to innovate the user experience. CVPR provided OPPO with an excellent platform to discuss and exchange ideas with global partners on the latest advances in computer vision and other AI-related fields.
OPPO’s AI research progress
This year, OPPO’s Research Institute and various R&D teams received significant recognition with 11 papers selected for presentation at CVPR. These studies cover key areas such as image restoration, digital human generation, video segmentation, dynamic motion capture, dynamic scene image synthesis, multi-view 3D object detection, and 3D rendering, highlighting some of the most innovative areas in the field of AI phones.
In the field of virtual character creation, OPPO has introduced a revolutionary approach with UltrAvatar, a method for creating controllable 3D virtual avatars that can be manipulated. Using a diffuse color extraction model and an authenticity-driven texture diffusion model, UltrAvatar can generate controllable 3D virtual characters from a single image or line of text. These virtual figures look indistinguishable from real people, achieving a level of detail that captures every inch of skin color and texture on a person’s face, greatly enhancing the realism of virtual interactions.
Another notable contribution, UniVS, introduces the industry’s first universal large-scale model for video segmentation. This model has the unique ability to address a variety of video segmentation needs, allowing all objects within a given category to be segmented and responding to textual descriptions for targeted segmentation.
These selected works not only represent cutting-edge achievements in the field of artificial intelligence, but also position OPPO as a leader advancing the mobile industry into the era of artificial intelligence.
In addition, the CVPR 2024 NTIRE RAIM (“Recovering Any Image Model”) competition, jointly organized by the Y-Lab of OPPO Research Institute and the Visual Computing Lab of the Hong Kong Polytechnic University, was a huge success. This competition encouraged researchers to solve practical real-world problems, develop methods for accurate, high-quality image restoration, and promote the application of these technologies in industry. More than 200 teams from around the world took part in the NTIRE RAIM competition. And in the end, the MiAlgo team’s research was the winner, taking first place in the competition.
Going forward, OPPO will actively support the AI development ecosystem by fostering close collaboration between academia and industry to bridge the gap between the two.
Find X7 Ultra debut
The Find X7 Ultra made its debut at CVPR, sparking interest in AI phone innovation.
OPPO’s flagship AI phone, the Find X7 Ultra, made its debut at the conference, quickly becoming a sensation at the event. In addition, Liang Jie, Senior Engineer of Image Processing Algorithms at Y-Lab, OPPO Research Institute, was invited to present a talk on OPPO’s image processing technology at the conference. The powerful AI image processing capabilities of Find X7 Ultra and OPPO’s innovative image generation applications have attracted the attention of both the academic and industrial AI development community.
AI functions
This year marks OPPO’s official entry into the era of AI phones: Find X7 Ultra is OPPO’s most advanced AI phone to date. It boasts the OPPO HyperTone Image Engine, the most advanced computer photography system currently available on the market. This algorithm allows OPPO Find X7 Ultra to create exceptional photos with minimal computing power, turning every shot into a masterpiece.
The OPPO AI Eraser feature introduced with the Find X7 Ultra demonstrated to CVPR attendees OPPO’s ability to seamlessly integrate AI technology into consumer devices to create a user-friendly interface. OPPO AI Eraser allows users to easily remove unwanted elements from an image and generate new content with a single tap. Since its introduction, this feature has revolutionized photo editing, with an average of 15 uses per day per user, making it one of the most popular features.
Other areas of AI technology
OPPO also demonstrated several research features on real demo devices at CVPR, highlighting the future direction of OPPO’s AI technology. Among them was the Style Transfer Demo, which utilizes ShaderNN’s lightweight deep learning architecture. ShaderNN is the first AI engine to support the full graphics stack (OpenGL and Vulkan) for real-time graphics and image processing in mobile applications, demonstrating OPPO’s strength in AI-enabled graphics generation on devices.
Democratizing phones with AI
At the beginning of the year, OPPO strategically established the OPPO AI Center, focusing the company’s resources on expanding the capabilities of artificial intelligence. This initiative has not only unified and strengthened OPPO’s research in various AI fields, but also aims to create the best AI phones. It has strengthened OPPO’s technical base in the AI phone sector and showcased impressive research results at this year’s CVPR conference.
This year’s CVPR conference also marks OPPO’s continued presence at leading international AI industry forums, after participating in events such as MWC, Google Cloud Next ’24, and Microsoft Build. OPPO is actively working with global partners to advance the development of AI phones.
In early June, OPPO officially announced its strategy to make AI phones accessible to everyone, aiming to equip approximately 50 million OPPO smartphones with generative AI features by the end of the year, covering all OPPO smartphone lines. This move underscores OPPO’s long-term commitment and strategic determination to strengthen the integration of AI into smartphones.
Going forward, OPPO’s innovative research findings, combined with cutting-edge technologies from partners such as Google, Microsoft, and MediaTek, will further advance OPPO’s mission to lead the global movement to democratize AI phones.
The post OPPO presents research in the field of AI first appeared on HiTechExpert.top.