A method for reconstructing photorealistic, animatable head avatars at speeds sufficient for on-the-fly reconstruction. Unlike prior approaches that utilize linear bases from 3D morphable models (3DMM) to model Gaussian blendshapes, our method maps tracked 3DMM parameters into reduced blendshape weights with an MLP, leading to a compact set of blendshape bases.
Programmable Optics for LiDAR and 3D Sensing: How Lumotive’s LCM is Changing the Game
For decades, LiDAR and 3D sensing systems have relied on mechanical mirrors and bulky optics to direct light and measure distance. But at CES 2025, Lumotive unveiled a breakthrough—a semiconductor-based programmable optic that removes the need for moving parts altogether.
The Problem with Traditional LiDAR and Optical Systems
LiDAR and 3D sensing systems work by sending out light and measuring when it returns, creating a precise depth map of the environment. However, traditional systems have relied on physically moving mirrors and lenses, which introduce several limitations:
Size and weight – Bulky components make integration difficult.
Complexity – Mechanical parts are prone to failure and expensive to produce.
Speed limitations – Physical movement slows down scanning and responsiveness.
To bring high-resolution depth sensing to wearables, smart devices, and autonomous systems, a new approach is needed.
Enter the Light Control Metasurface (LCM)
Lumotive’s Light Control Metasurface (LCM) replaces mechanical mirrors with a semiconductor-based optical chip. This allows LiDAR and 3D sensing systems to steer light electronically, just like a processor manages data. The advantages are game-changing:
No moving parts – Increased durability and reliability
Ultra-compact form factor – Fits into small devices and wearables
Real-time reconfigurability – Optics can adapt instantly to changing environments
Energy-efficient scanning – Focuses on relevant areas, saving power
How Does it Work?
LCM technology works by controlling how light is directed using programmable metasurfaces. Unlike traditional optics that require physical movement, Lumotive’s approach enables light to be redirected with software-controlled precision.
This means:
No mechanical delays – Everything happens at electronic speeds.
AI-enhanced tracking – The sensor can focus only on relevant objects.
Scalability – The same technology can be adapted for industrial, automotive, AR/VR, and smart city applications.
Live Demo: Real-Time 3D Sensing
At CES 2025, Lumotive showcased how their LCM-enabled sensor can scan a room in real time, creating an instant 3D point cloud. Unlike traditional LiDAR, which has a fixed scan pattern, this system can dynamically adjust to track people, objects, and even gestures on the fly.
This is a huge leap forward for AI-powered perception systems, allowing cameras and sensors to interpret their environment more intelligently than ever before.
Who Needs This Technology?
Lumotive’s programmable optics have the potential to disrupt multiple industries, including:
Automotive – Advanced LiDAR for autonomous vehicles
Industrial automation – Precision 3D scanning for robotics and smart factories
Smart cities – Real-time monitoring of public spaces
AR/VR/XR – Depth-aware tracking for immersive experiences
The Future of 3D Sensing Starts Here
Lumotive’s Light Control Metasurface represents a fundamental shift in how we think about optics and 3D sensing. By bringing programmability to light steering, it opens up new possibilities for faster, smarter, and more efficient depth-sensing technologies.
With traditional LiDAR now facing a serious challenge, the question is: Who will be the first to integrate programmable optics into their designs?
Avat3r takes 4 input images of a person’s face and generates an animatable 3D head avatar in a single forward pass. The resulting 3D head representation can be animated at interactive rates. The entire creation process of the 3D avatar, from taking 4 smartphone pictures to the final result, can be executed within minutes.
SPAR3D is a fast single-image 3D reconstructor with intermediate point cloud generation, which allows for interactive user edits and achieves state-of-the-art performance.
An efficient differentiable mesh-based method that can effectively handle complex 2D and 3D shapes. For instance, it can be used for reconstructing complex shapes from point clouds and multi-view images.
Odyssey and a Pixar co-founder, Ed Catmull (also an investor), just dropped Explorer, a revolutionary 3D world generator that turns any image into an editable 3D world.
A novel approach that leverages momentum-based self-distillation to promote consistency and accuracy across the blocks while decoupling the number of blocks from the physical GPU count.
• Slashes file sizes by 90% (250MB → 25MB) with virtually zero quality loss
• Lightning-fast uploads/downloads, especially on mobile
• Dramatically reduced memory footprint
• Enables real-time processing right on your phone
Tech breakthrough:
• Smart compression of position, rotation, color & scale data
• Column-based organization for maximum efficiency
• Innovative fixed-point quantization & log encoding
DISCLAIMER – Links and images on this website may be protected by the respective owners’ copyright. All data submitted by users through this site shall be treated as freely available to share.