Free Information Center
ROOTS is a dataset designed for research in environmental sound recognition and audio event detection. It comprises diverse audio recordings of natural and urban soundscapes aimed at advancing machine learning models in acoustic scene analysis.
A neural processing unit (NPU) is a specialized microprocessor designed to accelerate artificial intelligence (AI) and machine learning (ML) tasks. It optimizes the execution of neural network computations, enabling efficient processing of AI workloads in devices ranging from smartphones to data centers.
RWKV is a novel type of neural network architecture that combines recurrent neural network properties with transformer-level performance. It aims to provide efficient sequence modeling with scalability similar to transformers, while maintaining the benefits of recurrent structures.
AlphaGo is a computer program developed by DeepMind Technologies to play the board game Go. It was the first AI to defeat a professional human Go player, marking a significant milestone in artificial intelligence research.
Hyena operators are a class of sequence modeling techniques designed to efficiently process long-range dependencies in sequential data. They employ specialized structured state-space models to achieve scalability and performance improvements over traditional methods like transformers.
Ablation in neural network interpretability involves systematically removing components to assess their impact on model performance, aiding in understanding model decisions.
Neuralangelo is a 3D reconstruction technology that utilizes neural networks to generate detailed three-dimensional models from images or video data. It leverages advances in artificial intelligence to create accurate, textured models useful for various applications including cultural heritage preservation and virtual reality.
Hubert Dreyfus was an influential American philosopher known for his work in existentialism, phenomenology, and the philosophy of mind.
Sub-symbolic AI refers to a category of artificial intelligence that operates without explicit symbolic representation of knowledge, relying instead on connectionist or statistical methods.
Stochastic weight averaging (SWA) is an optimization technique used in training deep neural networks. It involves averaging multiple sets of weights collected at different points during the training process to improve generalization and model performance.