360° imaging
Drawing on our experience developing cameras such as RICOH THETA, we design capture and transmission from the camera side — through to live 360° streaming and continuous, long-running operation.
Spatial computing × 360° imaging
Capture reality.
Understand space.
Walk through a space with a 360° camera — that is all it takes. The space becomes a point cloud, then a photorealistic 3D scene, with every camera position on record. Building on years of camera engineering and image processing, we develop technology that captures the real world and makes sense of it.
01Why Infomorph
Give AI eyes, ears, and the field.
AI has become remarkably smart. But it still can't walk the site, look around and check for itself. We capture the field through photos, video, audio and sensors, turn it into data AI can work with, and help you decide which technology to use and how far to take it. That's where we come in.
Put your field data to work.
We collect and clean up photos, video, audio and sensor data, and shape it so AI can read it. With a background in camera engineering, we design from the moment of capture.
A technical eye, on your side.
From choosing technologies to quick technical validation and working with development vendors, we act as your tech lead. We follow new research every week, so decisions are made knowing the options.
AI that holds up in specialist work.
Generative AI can produce something plausible in minutes. Making it dependable in real work takes domain knowledge, data and validation. We team up with domain experts to build AI that goes deep into the work.
02What we do
Capture. Reconstruct. Locate. Interpret.
We work across the entire spatial pipeline — from apps running inside the camera to cloud processing and AI. Owning every stage lets us design systems end to end.
3D reconstruction & spatial understanding
From video of a room or building, we automatically generate point-cloud 3D maps and textured models — turning spatial geometry, and how it changes, into usable data.
Visual localization
We estimate where each image was taken within a 3D space — indoors, without GPS — including photos captured separately.
AI analysis
We read the state of things from images, audio, sensor data and biosignals, then analyze and visualize it — applying machine learning, deep learning or generative AI, whichever fits the problem.
03Products
From our lab to the field.
Technology only matters once people use it — so we turn our research into products.
View all productsInfomorph 3D
3D from video
Builds 3D maps from video and estimates camera positions — for automated spatial mapping and robotics.
TwinMaker (opens in a new tab)
Virtual tour cloud
Upload 360° video and a floor plan to generate a plan-linked street view of your site — for construction records and progress tracking.
Infomorph Live
360° live streaming
Streams live 360° video from RICOH THETA to a web browser — for remote site checks and virtual tours.
Spatial Computing Weekly (opens in a new tab)
Weekly research digest
An LLM-curated digest of notable spatial computing papers, updated every Sunday morning.
04Research
Following the research frontier, week by week.
New methods in spatial computing appear constantly. We track the latest research and bring what works into our own development.
Spatial Computing Weekly is that habit made public — a weekly digest of notable papers, with rankings, citation networks and a trend dashboard.
Research areas we follow
- Gaussian Splatting
- 360° panoramas
- Depth estimation
- 3D reconstruction & NeRF
- Point clouds & LiDAR
- VGGT
05News
Recent activity
We present our technology at exhibitions and talks. New work is announced in our news.
We have redesigned our website
We have reorganized our work around three areas and made our products, research and activities easier to find. The English site now mirrors the Japanese one.
Launching Spatial Computing Weekly, a weekly digest of new papers
We launched Spatial Computing Weekly, an LLM-curated weekly roundup of notable spatial computing papers.
Talk at JASVE Policy Committee & Tsunagu-Wa Session: Shintoshin Evening Salon
At the Tsunagu-Wa Session: Shintoshin Evening Salon (Sep 25, 2025), we introduce TwinMaker in a talk on easy-to-use localization technology.
Contact
What would your site tell you as data?
The earlier the question, the better — “Is this even possible?” is a great place to start. From technical validation to productizing a prototype and building it into your product, we will work it out with you.
Not ready to talk yet? Follow along