Product Design, Manufacturing & Innovation Resources

ディープラーニング

ニューラルネットワークDeep learning trains neural networks with many layers on large datasets, letting the network build its own features instead of relying on hand-written ones. It needs data, compute and careful validation. It gives no guarantee outside the distribution it was trained on.

New to the vocabulary? Begin with the article on the most important AI terms, which separates machine learning, deep learning, large language models and generative AI.

Everything else on this page is current research, digested from patents and journals. Computer vision and 3D SLAM serve robotics and augmented reality. Small language models target devices with little memory. Bioinformatics and cheminformatics apply the same networks to sequences and molecules, and brain-computer interfaces use them for signal decoding. Hardware constraints are discussed under edge AI inference.

10年目 周年

フルサイズの画像とダウンロードは、登録会員のみが100%無料で利用できます。