Strength of gender biases in AI images varies across languages
Peer-Reviewed Publication
Updates every hour. Last Updated: 16-Dec-2025 20:11 ET (17-Dec-2025 01:11 GMT/UTC)
Researchers at the Technical University of Munich (TUM) and TU Darmstadt have studied how text-to-image generators deal with gender stereotypes in various languages. The results show that the models not only reflect gender biases, but also amplify them. The direction and strength of the distortion depends on the language in question.
How can traditional coal-fired power plants adapt to the fluctuating nature of renewable energy? A new pre-gasification burner technology offers a solution, enhancing flexibility and stability even at ultra-low loads. Learn how this innovative approach could transform power plant operations and support a more sustainable energy future.
As the demand for high-quality, healthy solid-state lighting (SSL) grows, violet-light-excited full-spectrum lighting has emerged as a promising solution—it avoids blue light hazards and mimics natural sunlight. However, the critical yellow luminescent materials for this scheme are extremely scarce, plagued by low violet-light absorption and poor photoluminescent quantum yield (PLQY). To address this gap, a research team developed glass network engineering for the B2O3-BaO-Sc2O3 system, successfully fabricating violet-light-excitable yellow-emitting Ba2Sc2B4O11 (BSB):Ce3+ glass ceramics (GCs) with a record PLQY of 95.0% and superior thermal, moisture, and irradiation stability. By optimizing the [BO3]/[BO4] ratio, the team promoted heterogeneous nucleation during in-situ crystallization, forming well-crystallized BSB nanocrystals (NCs) in the glass matrix. This advancement enabled the construction of LED/LD-driven full-spectrum light sources with a color rendering index (CRI) exceeding 93, accelerating the development of sun-like lighting technology.
You won't find quantum computers in your local electronics store, but in the future, these superfast computers will probably become more common. Researchers from the Norwegian University of Science and Technology are bringing this future a small step closer.
A first-of-its-kind study explored whether more accessible technologies – foot-mounted wearable sensors and a 3D depth camera – could accurately measure how people walk, offering a practical alternative to traditional gait analysis tools. Gait, or walking pattern, is a key health indicator used to detect fall risk, monitor rehabilitation and identify early signs of conditions like Parkinson’s and Alzheimer’s. Traditional systems like the Zeno™ Walkway, the gold standard for gait analysis, are accurate but expensive, bulky and not easily used outside of lab settings.