|
|
|
|
|
Deep Learning Method Produces Holograms Instantly
Technology Briefing |
|
TranscriptDespite years of hype, 3D virtual reality headsets, TVs, and computer screens have not conquered the market. One reason is that 3D, especially when used for virtual reality makes users feel sick. Nausea and eye strain result because VR creates an illusion of 3D viewing although the user is, in fact, staring at a fixed distance, 2D display. The solution for achieving widespread consumer embrace of 3D visualization could lie in reviving holograms, a 60-year-old technology remade for the digital world. Holograms deliver an exceptional representation of the 3D world around us which is also beautiful. They also offer a shifting perspective based on the viewer's position, and they allow the eye to adjust the focal depth to alternately focus on foreground and background. Researchers have long sought to make computer-generated holograms, but the process has traditionally required a supercomputer to churn through physics simulations, which is time-consuming and can yield less-than-photorealistic results. But now, MIT researchers have developed a new way to produce holograms almost instantly and the new method based on deep learning is so efficient that it can run on a laptop in the blink of an eye. People have assumed that with existing consumer-grade hardware, it was impossible to do real-time 3D holography computations. And this has led to a serious paradox: experts have often said that "commercially available holographic displays will be available in 10 years." Yet, they've been making that same statement for six decades now. The MIT team believes the new approach, which they call "tensor holography," will finally bring that elusive "10-year goal" within reach. And this advance could quickly fuel a spillover of holography into fields like VR and 3D printing. A typical lens-based photograph encodes the brightness of each light wave striking a surface. Therefore, a photo can faithfully reproduce a scene's colors, but it ultimately yields only a flat image. In contrast, a hologram encodes both the brightness and phase of each light wave. That combination delivers a truer depiction of a scene's parallax and depth. So, while a photograph of Monet's "Water Lilies" can highlight the paintings' color palate, a hologram can bring the work to life, rendering the unique 3D texture of each brushstroke. Unfortunately, holograms have been a serious challenge to make and share. Computer-generated holography sidesteps the challenges associated with optically generated holograms. But, until now, that process has been a computational slog. Because each point in the scene has a different depth, you can't apply the same operations for all of them. That increases the complexity significantly. And directing a cluster of supercomputers to run these physics-based simulations could take seconds or minutes for a single holographic image. Furthermore, existing holography algorithms don't model occlusion with photorealistic precision. So, the MIT team took a different approach: letting the computer teach physics to itself. They used deep learning to accelerate computer-generated holography, allowing for real-time hologram generation. The team designed a convolutional neural network - a processing technique that uses a chain of trainable tensors to roughly mimic how humans process visual information. Training a neural network typically requires a large, high-quality dataset, which didn't previously exist for 3D holograms. So, the team built a custom database of 4,000 pairs of computer-generated images. Each pair-matched a picture - including color and depth information for each pixel - with its corresponding hologram. To create the holograms in the new database, the researchers used scenes with complex and variable shapes and colors, with the depth of pixels distributed evenly from the background to the foreground, and with a new set of physics-based calculations to handle occlusion. That approach resulted in photorealistic training data. Next, the algorithm got to work. By learning from each image pair, the tensor network tweaked the parameters of its own calculations, successively enhancing its ability to create holograms. The fully optimized network operated orders of magnitude faster than physics-based calculations. The resulting efficiency surprises even the team members. They are amazed at how well it performs. In mere milliseconds, tensor holography crafts holograms from images with depth information calculated from a multicamera setup or LiDAR sensors. This advance paves the way for real-time 3D holography. What's more, the compact tensor network requires less than 1 MB of memory. In fact, it's negligible, considering the tens and hundreds of gigabytes available on the latest cell phone. Looking ahead, real-time 3D holography could enhance a slew of systems, from VR to 3D printing. The team says the new system could help immerse virtual reality in more realistic scenery while eliminating eye strain and other side effects of long-term VR use. And the technology could be easily deployed on displays that modulate the phase of light waves. Currently, most affordable consumer-grade displays modulate only brightness, though the cost of phase-modulating displays would fall if widely adopted. Three-dimensional holography could also boost the development of volumetric 3D printing, the researchers say. This technology could prove faster and more precise than traditional layer-by-layer 3D printing since volumetric 3D printing allows for the simultaneous projection of the entire 3D pattern. Other applications include microscopy, visualization of medical data, and the design of surfaces with unique optical properties. Digital holography represents a considerable leap that could completely change people's attitudes toward 3D. And it looks like neural networks were born for this task. |
|
Comments
|
|
|
|
|
|
|