HuggingFace's platform is a resource for developers and researchers working in natural language processing (NLP) and machine learning, offering insights into NLP models, tools, and datasets. Through articles, tutorials, and open-source projects, HuggingFace offers insights into state-of-the-art NLP techniques, transformer architectures, and transfer learning methods. Developers can learn about using pre-trained models, fine-tuning strategies, and deploying NLP applications with HuggingFace's libraries and APIs.

Hugging Face

NVIDIA and Siemens Healthineers have developed NV-Raw2Insights-US, an AI model that learns directly from raw ultrasound channel data rather than reconstructed images. The model estimates patient-specific speed-of-sound maps to enable adaptive image focusing in real time, replacing traditional hand-engineered beamforming pipelines that make simplifying physics assumptions. Deployment uses NVIDIA Holoscan Sensor Bridge with an FPGA to stream raw data from existing ultrasound scanners via DisplayPort, then runs inference on Blackwell-class GPUs. The architecture is modular, allowing future AI models to be integrated without hardware changes. Model weights, dataset, and code are publicly available on GitHub and Hugging Face.

Adaptive Ultrasound Imaging with Physics-Informed NV-Raw2Insights-US AI