#Neural Architecture Search

The DAMO-YOLO project, developed by the TinyML Team from Alibaba's DAMO Lab, enhances object detection beyond the YOLO series through innovative technologies like NAS backbones and efficient RepGFPN. It delivers powerful models and high-efficiency training tools, applicable to diverse scenarios including human and safety equipment detection. Updated models exhibit significant gains in real-time performance on both CPUs and GPUs. This project is resourceful for applications requiring precise and rapid object detection capabilities.

Explore a curated list of the latest Automated Machine Learning (AutoML) research papers, articles, and projects. Gain insights into vital AutoML processes such as data preprocessing, feature selection, hyperparameter optimization, and performance prediction. Reflecting on the growing corporate interest, it illustrates how major companies and startups are developing AutoML systems, showcasing their potential to rival human expertise. An essential resource for those seeking to delve into or contribute to the evolving AutoML landscape.

This project offers a detailed exploration of dataset distillation, focusing on transforming large datasets into smaller, highly efficient versions. It underscores the importance of distilled datasets in fields such as continual learning, privacy preservation, and neural architecture search. Since its introduction in 2018, the project has captured significant advancements, featuring crucial research efforts, methods like gradient and distribution matching, and practical use cases. Maintained by experts Guang Li, Bo Zhao, and Tongzhou Wang, this compilation acts as an essential resource for those interested in dataset optimization and its diverse applications.