Healthcare. Where we began. Imaging.

In 2000, 3dMD helped transition the health profession from 2D photography to anatomically precise, ultrafast non-invasive 3D dense-surface imaging. In 2013, 3dMD led a second transition… from static 3D dense-surface imaging to dynamic 3D/4D dense-surface imaging systems that capture and reconstruct progressive sequences of a patient’s anatomical mobility and dynamic shape change. Twenty-five years later, 3dMD’s software-driven systems are still considered the imaging modality standard for near ground truth 3D-shape information for anthropometric analysis across medical and dental communities worldwide… a position backed by hundreds of peer-reviewed research publications produced by the global 3dMD customer community.

3dMD images continue to support clinical teams worldwide in assessing surgical intervention, objectively measuring morphological change throughout the treatment cycle, and evaluating outcomes. From complex long-term pediatric treatment cycles such as cleft lip and palate, through to craniofacial, oral maxillofacial (OMS), reconstructive, and trauma surgery, printed prosthetics, and orthopedics, 3dMD has built its reputation where the 3D-shape accuracy demands are highest.

Clinical teams are no longer limited to single static 3D scans of disparate expressions, poses, and postures. 3dMD’s software-driven, dynamic-3D/4D imaging systems capture progressive sequences of dense-surface images throughout the treatment cycle… enabling clinical teams to select the optimal 3dMD image from the sequence or analyze the full dynamic recording of a patient’s anatomical function, movement, pose, posture, communication, and expression over time. For forward thinking clinical teams, 3dMD’s downstream 3dMDtempus software enables users to automatically landmark, track, measure, and analyze facial movement, expression, function, and speech with audio sync.

Computer Vision | Perception. Training.

Computer Vision aims to automate programs that replicate the human visual system. To do that realistically and reliably, it needs real-world 3D/4D data… elevating what synthetic approximations, sparse markers, and 2D video alone cannot achieve.

3dMD is at its core a software-driven computer vision application. That foundation makes 3dMD systems a natural fit for customer teams building their own computer vision applications. 3dMD provides progressive sequences of real-world 3D dense-surface images with a predictable, consistent near ground truth 3D-shape reconstruction… giving research teams a reliable base of temporally consistent, spatially dense, and anatomically accurate frames to establish deep learning resources that document and interpret quantitative human behavior.

Sophisticated body and hand models such as SMPL and MANO… developed by the world-class 3dMD customer team at Max Planck Institute for Intelligent Systems, Perceiving Systems, in Tübingen… are built with 3dMD’s real-world dense-surface 3D/4D image sequences. Meta Reality Labs recently released PALM… a large-scale hand dataset built entirely on 3dMD, comprising 13k registered 3dMD hand scans from 263 subjects with 90k calibrated 3dMD multiview RGB images, each with corresponding MANO registrations.

Whether the goal is contactless device interaction or teaching systems to interpret and predict implicit human intent, 3dMD customer teams are building Real-World 3dMD Humans in motion… the most accurate training foundation available.

Artificial Intelligence | Machine Learning. Training.

Simply training AI and Machine Learning models on synthetic data, sparse markers, and 2D video has limits. Real-world human behavior captured in 3D/4D with all of the conventions from the complex and subtle to the mundane… is what separates realistic, reliable dynamic spatial models from low fidelity ones that fall short in digitizing the real world.

3dMD customer teams are building vast, diverse real-world databases of dense-surface 3D/4D human subjects performing natural actions… gesturing and communicating with their face, body, limbs, and hands… captured at 10/30/60/120fps throughout the entire subject performance. The result is rich, near ground truth 3D-shape data sequences reconstructed by 3dMD’s Active Stereo Photogrammetry software algorithms… feeding deep learning models with the temporal consistency, spatial density, and anatomical 3D-shape accuracy that synthetic datasets cannot replicate.

Because of the richness of real-world 3dMD data, it has persistent value… it can be re-mined to explore further detail, flow, and structure as AI models develop and demands evolve.

Humanoid Robotics. Training.

Humanoid robots require more than 2D video, sparse markers, and synthetic approximations of human motion for life-like imitation learning. To achieve human-level dexterity, perception, and interaction, humanoid training platforms need to assimilate the complexities and nuances of Real-World 3D/4D Human behavior and manipulation intelligence. The uncanny valley is the challenge… a humanoid robot that moves, gestures, or expresses itself with even subtle inaccuracy in a social environment immediately triggers an instinctive unease in the people it is designed to work alongside. Bridging that gap requires the most accurate real-world human data available.

3dMD, a proven software-driven imaging modality in the clinical sector, delivers the world’s most accurate dynamic 3D/4D human capture and reconstruction technology… making the reconstructed output a natural and fundamental data source for deep training next-generation adaptable humanoid robots.

3dMD customer teams use real-world 3D/4D capture across the full human form… body, hands, and face… to train the humanoid software brain where the real intelligence exists. High-resolution 360-degree body data trains natural locomotion, balance, spatial reasoning, gesture, and gait. Detailed hand data captures real grip dynamics, precision tool use, delicate interaction, and bimanual coordination across progressive 3D/4D sequences… delivering the temporal consistency, spatial density, and anatomical ground truth essential for dexterous hand control. Before a humanoid robot hand can be teleoperated or trained, it needs to be designed… and achieving human-like dexterity requires understanding the actual 3D surface geometry, soft tissue deformation, and skin compliance of real human hands across diverse populations. Motion capture gloves record joint angles and finger kinematics for teleoperation demonstrations, but they do not capture the real 3D surface of the hand… and the glove itself physically alters the surface it is measuring. 3dMD captures the ground truth surface geometry needed to design robot hands that replicate human interaction, train AI models that generalise across diverse hand morphologies, and validate that the final design interacts with objects the way a real human hand does. High-fidelity face data captures expressions, speech, eye gaze, and subtle muscle movements critical for human-robot interaction.

Real-world 3dMD data reduces AI training errors, fills critical gaps in synthetic datasets, and closes the sim-to-real divide. Because it is captured and reconstructed with medical-grade 3D accuracy across diverse ages, ethnic backgrounds, genders, and body types, it has persistent value… re-mined as humanoid platforms evolve.

Real-world 3dMD data goes beyond teaching humanoids how to move. It captures the subtleties of when, why, and in what way… within an interactive work and social context.

Meta Reality Labs recently released PALM… 13k registered 3dMD hand scans from 263 subjects with 90k calibrated 3dMD multiview RGB images, each with corresponding MANO registrations. The most widely used AI hand models are built on 3dMD data, not glove data.

Wearable Technology. Wearing.

Wearable devices need to fit, function, and move with the human body. 3dMD customer teams use real-world 3D/4D human data to ergonomically design devices to actual subject body types and calibrate wear against natural movement and daily activity.

Wearable design traditionally relies on static-3D scans… optimizing fit for a head, hand, body shape that is standing still. But people move. Specific to the head, people talk, chew, turn, look down, and express. A device that fits perfectly on a static 3D scan can shift, pinch, or lose optical alignment during natural daily use. 3dMD’s dynamic-3D/4D capture capability enables design teams to evaluate fit, stability, and comfort throughout real subject movement… transitioning wearable design from fitting a mannequin to fitting a living person. Capturing subjects wearing and not wearing a device with the same system platform enables direct comparison of contact areas, interference points, and as-worn position across users.

The progressive sequence of 3dMD dense-surface images captures a subject’s full dynamic movement while wearing a device across diverse body types, ages, and ethnic backgrounds… identifying design refinements that improve performance, comfort, and user experience for the full range of the intended population. A single 3dMD image in the sequence can be equally valuable… pinpointing a specific pose or position where fit breaks down.

Wearable technology is a natural evolution of 3dMD’s early work supporting customer design teams in advanced prosthetics, orthotics, and performance sportswear… now applied across AR/VR eyewear, head-mounted displays, hearables, smartwatches, fitness trackers, smart clothing, and Personal Protection Equipment.

Leading AR companies such as Magic Leap and many others use 3dMD’s anthropometric data as the foundation of their device design process… capturing thousands of subjects across age, gender, and ethnic background to design AR devices around real human variability rather than a single average fit model. As Magic Leap describes it, simulation is just the start of the design process… empirical fit data from actual users is what bridges the gap between virtual models and comfort-optimized physical devices.

Human Factors | Anthropometric | Size and Fit. Wearing.

Traditional sizing systems are built around a single average fit model… ignoring the real-world variety of body shapes within any given size category. 3dMD customer teams are actively addressing this by capturing real-world body type variations across size, age, ethnic background, and mobility categories and population samples with meaningful ethnic and age diversity.

For head-worn devices, the challenge extends beyond static fit. Human factors teams evaluating eyewear, headsets, and helmets need to understand how diverse head shapes interact with device geometry across the full range of natural movement… talking, chewing, turning, and looking down. 3dMDhead system configurations from 5-12 MCUs provide the line-of-sight coverage needed to capture complex anatomy including the ear region, eye region, and nasal bridge where device contact and comfort are most critical.

3dMD systems deliver precise, real-world anatomical 3D shape and size data efficiently and without physical contact… optimizing the subject workflow from recruitment through to data output. Whether the application is safety gear, footwear, performance sportswear, uniforms, seating, or on-demand apparel production, 3dMD captures the individual’s full dynamics and anatomical 3D geometry at the accuracy levels that legacy sizing systems, traditional measurement methods, static-3D scanning, and parametric model-dependent approaches simply cannot match.

XR | AR | VR. Training.

Convincing virtual environments depend on one thing… realistic, expressive 3D/4D human avatars that interact naturally with their environment, objects, and other humans. The uncanny valley is the challenge. Avatars that fall short of true human fidelity trigger an instinctive unease in the viewer… undermining the entire experience. Synthetic approximations cannot bridge that gap. Real-world 3D/4D data can.

3dMD customer teams are building diverse Real-World 3dMD Human performance, interaction, and emotion databases to train avatars at the fidelity that VR demands. Separately, teams are capturing progressive sequences of real-world hand gesture and finger articulation to develop intuitive UI control systems that enable users to reach out and interact naturally with objects in the AR environment.

Real-world 3dMD data delivers what synthetic datasets cannot… the full complexity of human movement, expression, interaction, and nuance captured across temporally consistent, spatially dense progressive sequences with near ground truth 3D-shape accuracy.

Genetics | Biomarkers. Imaging.

External biomarkers… visible wellness changes projected onto the surface of the human body… are an increasingly important tool in a preventative care approach. 3dMD customer teams in genetics and disease research are using real-world 3D surface imaging to identify, define, and study morphological anatomical structures that may serve as visual indicators of underlying conditions.

3dMD images enable research teams to establish and interpret their own biomarker sets with near ground truth 3D-shape accuracy… supporting early detection research and more effective study outcomes. Because 3dMD progressive image sequence data is persistent and can be re-mined as research evolves… tracking morphological change over time and across populations… findings published today continue to support new discoveries tomorrow.

To date this work has contributed to hundreds of peer-reviewed publications spanning genetics, disease control, and preventative care.