feature pooling
[/ˈfiːtʃər ˈpuːlɪŋ/]
noun
agrupamento de características
1. A technique in machine learning and deep learning where features from a neural network layer are combined or aggregated to reduce dimensionality while preserving important information
Feature pooling is commonly used in convolutional neural networks to downsample feature maps after convolution operations.
O agrupamento de características é comumente usado em redes neurais convolucionais para reduzir o tamanho dos mapas de características após operações de convolução.
2. The process of combining multiple feature representations into a single, more compact representation through various aggregation methods such as max pooling, average pooling, or sum pooling
The model uses feature pooling to merge spatial information from overlapping regions of the input image.
O modelo usa agrupamento de características para mesclar informações espaciais de regiões sobrepostas da imagem de entrada.
This is a specialized technical term primarily used in artificial intelligence, machine learning, and computer vision fields. It is predominantly used in academic and professional contexts in both Brazil and Portugal. In Brazilian tech companies and universities, the term is often used alongside the English original, especially in technical documentation and research papers. The concept is fundamental to modern deep learning architectures and is taught in advanced computer science programs.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →