Connect with us
We’re experimenting with AI-generated content to help deliver information faster and more efficiently.
While we try to keep things accurate, this content is part of an ongoing experiment and may not always be reliable.
Please double-check important details — we’re not responsible for how the information is used.

Civil Engineering

New Hybrid AI Tool Generates High-Quality Images Faster Than State-of-the-Art Approaches

Researchers developed a hybrid AI approach that can generate realistic images with the same or better quality than state-of-the-art diffusion models, but that runs about nine times faster and uses fewer computational resources. The tool uses an autoregressive model to quickly capture the big picture and then a small diffusion model to refine the details of the image.

Avatar photo

Published

on

The ability to generate high-quality images rapidly is crucial for producing realistic simulated environments that can be used to train self-driving cars to avoid unpredictable hazards. However, current generative AI techniques have drawbacks. Diffusion models can create stunningly realistic images but are too slow and computationally intensive for many applications. On the other hand, autoregressive models, which power LLMs like ChatGPT, are much faster but produce poorer-quality images.

Researchers from MIT and NVIDIA have developed a new approach that combines the best of both methods. Their hybrid image-generation tool, known as HART (Hybrid Autoregressive Transformer), uses an autoregressive model to quickly capture the big picture and then a small diffusion model to refine the details of the image. This innovative approach enables HART to generate images that match or exceed the quality of state-of-the-art diffusion models but do so about nine times faster.

The generation process consumes fewer computational resources than typical diffusion models, allowing HART to run locally on a commercial laptop or smartphone. A user only needs to enter one natural language prompt into the HART interface to generate an image.

HART has a wide range of potential applications, including helping researchers train robots to complete complex real-world tasks and aiding designers in producing striking scenes for video games. If you are painting a landscape, and you just paint the entire canvas once, it might not look very good. But if you paint the big picture and then refine the image with smaller brush strokes, your painting could look a lot better. That is the basic idea with HART.

The researchers encountered challenges in effectively integrating the diffusion model to enhance the autoregressive model. They found that incorporating the diffusion model in the early stages of the autoregressive process resulted in an accumulation of errors. Instead, their final design of applying the diffusion model to predict only residual tokens as the final step significantly improved generation quality.

Their method uses a combination of an autoregressive transformer model with 700 million parameters and a lightweight diffusion model with 37 million parameters. This allows HART to generate images of the same quality as those created by a diffusion model with 2 billion parameters but do so about nine times faster, using about 31 percent less computation than state-of-the-art models.

The future applications of this technology are vast and exciting. In the future, one could interact with a unified vision-language generative model, perhaps by asking it to show the intermediate steps required to assemble a piece of furniture. The researchers want to go down this path and build vision-language models on top of the HART architecture, since HART is scalable and generalizable to multiple modalities. They also want to apply it for video generation and audio prediction tasks.

Civil Engineering

A Groundbreaking Magnetic Trick for Quantum Computing: Stabilizing Qubits with Exotic Materials

Researchers have unveiled a new quantum material that could make quantum computers much more stable by using magnetism to protect delicate qubits from environmental disturbances. Unlike traditional approaches that rely on rare spin-orbit interactions, this method uses magnetic interactions—common in many materials—to create robust topological excitations. Combined with a new computational tool for finding such materials, this breakthrough could pave the way for practical, disturbance-resistant quantum computers.

Avatar photo

Published

on

By

The article you provided was well-written, but I made some adjustments to improve clarity, structure, and style for general readers. Here’s the rewritten content:

A Groundbreaking Magnetic Trick for Quantum Computing: Stabilizing Qubits with Exotic Materials

Quantum computers have long been touted as revolutionaries in solving complex problems that conventional supercomputers can’t handle. However, their development has been hindered by one major challenge: qubits, the basic units of quantum computers, are extremely delicate and prone to losing their quantum states due to external disturbances.

Researchers from Chalmers University of Technology in Sweden and Aalto University and the University of Helsinki in Finland have now made a groundbreaking discovery that could change this. They’ve developed a new type of exotic quantum material that exhibits robust topological excitations, which are significantly more stable and resilient than other quantum states.

This breakthrough is an important step towards realising practical topological quantum computing by constructing stability directly into the material’s design. The researchers’ innovative approach uses magnetism as the key ingredient to achieve this effect, harnessing magnetic interactions to engineer robust topological excitations in a broader spectrum of materials.

“The advantage of our method is that magnetism exists naturally in many materials,” explains Guangze Chen, postdoctoral researcher in applied quantum physics at Chalmers and lead author of the study published in Physical Review Letters. “You can compare it to baking with everyday ingredients rather than using rare spices. This means that we can now search for topological properties in a much broader spectrum of materials, including those that have previously been overlooked.”

To accelerate the discovery of new materials with useful topological properties, the research team has also developed a new computational tool that can directly calculate how strongly a material exhibits topological behavior.

“Our hope is that this approach can help guide the discovery of many more exotic materials,” says Guangze Chen. “Ultimately, this can lead to next-generation quantum computer platforms, built on materials that are naturally resistant to the kind of disturbances that plague current systems.”

This magnetic trick has the potential to revolutionize the development of practical topological quantum computing and pave the way for next-generation quantum computer platforms. As researchers continue to explore and develop new exotic materials with robust topological excitations, we may finally see the dawn of a new era in quantum computing.

Continue Reading

Civil Engineering

AI Breakthrough in Fusion Reactor Design: Uncovering Hidden Safe Zones with HEAT-ML

Scientists have developed a lightning-fast AI tool called HEAT-ML that can spot hidden “safe zones” inside a fusion reactor where parts are protected from blistering plasma heat. Finding these areas, known as magnetic shadows, is key to keeping reactors running safely and moving fusion energy closer to reality.

Avatar photo

Published

on

By

The development of artificial intelligence (AI) in fusion research has taken a significant leap forward. A public-private partnership between Commonwealth Fusion Systems (CFS), the U.S. Department of Energy’s Princeton Plasma Physics Laboratory (PPPL), and Oak Ridge National Laboratory has led to the creation of HEAT-ML, an AI approach that rapidly finds and simulates “magnetic shadows” in fusion vessels: safe havens protected from intense heat plasma.

HEAT-ML uses a deep neural network to learn how to calculate shadow masks, which are 3D maps of specific areas on internal components shielded from direct heat. This AI surrogate was trained using a database of approximately 1,000 SPARC simulations and can now simulate the same calculations in mere milliseconds, as opposed to the previous 30 minutes.

The goal is to create software that significantly speeds up fusion system design and enables good decision-making during operations by adjusting plasma settings to prevent potential problems. HEAT-ML was specifically designed for a small part of the SPARC tokamak under construction by CFS but has the potential to be expanded to generalize the calculation of shadow masks for exhaust systems of any shape and size, as well as other plasma-facing components.

Researchers believe that this AI breakthrough could pave the way for faster fusion system design, enabling good decision-making during operations, and potentially leading to limitless amounts of electricity on Earth.

Continue Reading

Chemistry

Unlocking the Secrets of Atomic Motion: A Revolutionary Discovery at the Nanoscale

A pioneering team at the University of Maryland has captured the first-ever images of atomic thermal vibrations, unlocking an unseen world of motion within two-dimensional materials. Their innovative electron ptychography technique revealed elusive “moiré phasons,” a long-theorized phenomenon that governs heat, electronic behavior, and structural order at the atomic level. This discovery not only confirms decades-old theories but also provides a new lens for building the future of quantum computing, ultra-efficient electronics, and advanced nanosensors.

Avatar photo

Published

on

By

The study of atomic-scale phenomena has led researchers to a groundbreaking discovery that could reshape the design of quantum technologies and ultrathin electronics. Yichao Zhang, an assistant professor in the University of Maryland Department of Materials Science and Engineering, has developed an innovative technique called “electron ptychography” to directly image the thermal vibrations of individual atoms. This achievement was published in the journal Science on July 24.

Two-dimensional materials, which are sheet-like structures a few nanometers thick, have been explored as new components for next-generation quantum and electronic devices. A crucial feature of twisted two-dimensional materials is “moiré phasons,” essential to understanding their thermal conductivity, electronic behavior, and structural order. However, detecting moiré phasons experimentally had proven challenging, hindering further research in these revolutionary materials.

Zhang’s team overcame this challenge by employing electron ptychography, a technique that achieved the highest resolution documented (better than 15 picometers) and detected the blurring of individual atoms caused by thermal vibrations. This groundbreaking study revealed that spatially localized moiré phasons dominate thermal vibrations in twisted two-dimensional materials, fundamentally reshaping our understanding of their impact.

The breakthrough confirmed long-standing theoretical predictions of moiré phasons and demonstrated that electron ptychography can be used to map thermal vibrations with atomic precision for the first time. This achievement opens up new possibilities for exploring previously hidden physics in quantum materials.

“This is like decoding a hidden language of atomic motion,” said Zhang. “Electron ptychography lets us see these subtle vibrations directly. Now we have a powerful new method to explore previously hidden physics, which will accelerate discoveries in two-dimensional quantum materials.”

Zhang’s research team will next focus on resolving how thermal vibrations are affected by defects and interfaces in quantum and electronic materials. Controlling the thermal vibration behavior of these materials could enable the design of novel devices with tailored thermal, electronic, and optical properties – paving the way for advances in quantum computing, energy-efficient electronics, and nanoscale sensors.

Continue Reading

Trending