Frontline | Insta360 Hosts Smart Imaging Hackathon, Expanding from AI Editing to the Mobile Photographer Domain

Written by | Zhang Ziyi

Edited by | Yuan Silai

Recently, Insta360 held the 2026 Bold Maker Intelligent Imaging Challenge in Nanjing. More than 220 young makers completed prototype development within four days and three nights, and 15 teams advanced to the final to compete for 300,000 RMB in cash and 200,000 RMB in prizes.

During the competition, a team composed of students from Nanjing University tried to solve the "last 10 meters" problem for visually impaired people when traveling: when map navigation can only guide users to the vicinity of the destination, panoramic devices continue to observe the surrounding environment, AI identifies entrances and obstacles, and then prompts directions and distances through voice. Luo Jingyi, a member of the project, originally studied architecture. After interviewing visually impaired people, the team applied the architectural idea of connecting the "last kilometer" to the last more than ten meters of travel.

Other competing works are not only focused on photography: some people use intelligent imaging balls to collect images of pets' oral cavities, some use cameras to identify sign language and generate voice and display text in real time, and another team installed a panoramic camera on a self-developed underwater platform, completing assembly and underwater tests within 48 hours. All the above works point to one change — cameras are no longer just recording tools, but have begun to undertake environmental perception, understanding and feedback tasks.

This also echoes the AI technology route that Insta360 is adjusting.

Qi Lu, head of the Insta360 Research Institute and Chief Artificial Intelligence Officer, introduced during the event that Insta360 divides intelligent imaging into four links: "selection, shooting, editing, and transmission": enabling AI to participate in composition and scene interaction, improving image quality and sound processing, completing automatic editing, and exploring new ways of content sharing. As a result, AI is no longer just a single-point function in the camera, but is integrated into the complete link from scene selection to content dissemination.

36Kr learned that the planned route described above has been implemented in specific products. Kira, a multimodal voice assistant, can interact with users by combining the scenes seen by the lens; "Moment Pro" completes cloud-based automatic editing through general large models and PanoMind, a self-developed panoramic model of Insta360; "3D Time Warehouse" uses 3D reconstruction technology to convert panoramic videos into spatialized content.

Insta360 disclosed that the number of paid subscribers for its cloud storage subscription service Insta360+ increased by more than 124% year on year. Qi Lu said that automatic editing and AI services have become one of the main reasons for users to purchase devices.

Insta360's vision for AI is not limited to reducing post-creation costs. The company publicly launched the "Mobile Photographer" program for the first time, hoping that the device can observe the environment on its own, understand shooting targets, plan movement trajectories, and judge what is worth recording and when to press the shutter.

Qi Lu summarized this system as: AI is the brain, the lens is the eye, the drone or robot is the torso, and the motorized gimbal acts as the limbs. This concept was first proposed internally at Insta360 as early as six or seven years ago, and the development of embodied intelligence and large models has made its productization possible again.

Technically, Insta360 chooses to start from panoramic vision. Compared with ordinary cameras with limited field of view, panoramic devices can record 360-degree information at one time, reduce the information increment caused by rotating the lens, and provide more complete input for environment understanding and trajectory planning. Focusing on this direction, Insta360 has developed a panoramic depth model, a panoramic visual language dataset and a simulation platform, and is exploring ways to allow devices to independently decide how to move and how to shoot. The company said that relevant basic capabilities will be open to the research community in the form of papers, datasets and tools.

From panoramic cameras, action cameras to gimbal cameras and panoramic drones, Insta360 is expanding its boundaries from imaging hardware to software services and autonomous shooting systems.

This hackathon event provides another entry point: allowing external developers to use real-world problems to test what else panoramic technology and AI can do. Insta360 said that outstanding projects will have the opportunity to continue to be incubated and be considered in the R&D of next-generation devices, and participants may also become future employees.