„… der inneren Geometrie der Fläche ab, d. h. von der ersten Fundamentalform (bzw. dem metrischen Tensor …” „… lässt sich verallgemeinern auf Mannig faltigkeit beliebiger Dimension mit einem metrischen Tensor. Auf …”

„… oberen Flansch mit einem „Tensor-Revo“- Schrauber montiert, der an einem flexiblen Handlingarm …” „… konzentrieren. Bei der für Rollstar umgesetzten Lösung wurde in den Articulated Arm ein Tensor-Revo-Werkzeug …” „… der Fertigungsprozess optimiert, und zwar mit einem gesteuerten Hochmomentschrauber des Typs „Tensor …”

„… builds on Keras 2.2.4 [15] with TensorFlow 1.15.0.dev20190718 [16] [17] and TensorFlow probability 0.8.0 …” „… Optimization, 2018. [15] F. Chollet und others, Keras, 2015. [16] Martin Abadi und others, TensorFlow …” „… : Large-Scale Machine Learning on Heterogeneous Systems, 2015. [17] Magnus Erik Hvass Pedersen, TensorFlow …”

„… Tensoren konvertiert, und anschließend mehrere Bilder zu einem Batch zusammengefasst. Durch Batches werden …”

„… ]. Das gelernte prädiktive Modell zur Qualitätsvorhersage basiert auf Tensorflow [10] und wird in einer …” „… Material-Datenbank (Lieferant) Qualitäts- Vorhersage (Tensorflow) Optimierung (Apache Spark) Temperatureinstellung …” „… . "Tensorflow: A system for large-scale machine learning." In 12th USENIX Symposium on Operating Systems Design …”

„… vector, bias tensor β position angle CS cosine similarity concat ( ) concatenation operator, stacks …” „… tensors along the u3 dimension C marks the use of camera coordinates in [m] χ count of frequent classes …” „… optimization gMA moving average of the squared gradients h, h, H feature scalar, feature vector, feature tensor …”

„… bold lowercase letters, and matrices and tensors as bold uppercase letters. Accordingly, the mapping of …” „… , in any practical implementation H becomes a tensor with at least three dimensions (u1, u2, u3). This …” „… dimension, which is encoded in the third tensor coordinate u3. For the input images, this corresponds to the …”

„… through tensor concatenation. All bottlenecks layers compress the feature channel dimension to u3 = 10. 53 …” „… performance. Tensor concatenation along the u3 dimension can be used instead to merge the feature maps, which …” „… road segmentation ground truth and output feature map tensors of training sample j, yNL,j and hNL,j …”

„… . Tucker, V. Vasudevan, P. Warden, M. Wicke, Y. Yu, and X. Zheng. “TensorFlow: A System for Large-scale …” „… : https://cloud.google.com/ edge-tpu. Accessed: July 9th, 2020. [Goo19b] Google LLC. TensorFlow Lite. 2019. url …” „… : https://www.tensorflow. org/lite. Accessed: July 9th, 2020. [Gra14] B. Graham. “Fractional Max-Pooling”. In: arXiv …”

„… reconstruction the respective tensors for the offset localization output path, then the combined loss function is …”

„… processing models is largely based on the open source Tensorflow software framework [Aba+16] for the …” „… definition, training, and execution of deep neural network models. Here, the decisive advantage of Tensorflow …” „… development projects forked from Tensorflow [SJ19]. Mobile platform: test vehicle The examination of the image …”

„… https://github.com/tensorflow/models/blob/master/research/slim/nets/inception_v2.py . Accessed: September 21st, 2020 37 …”

„… are the fourth dimension of the input tensor instead of the third.Following the methodology of …” „… tensors. A one-hot encoded representation vector,and therefore number of channels of an input tensor can …” „… become verylarge, resulting in very sparse input tensors. Such a tensor is difficult totrain on for an NN …”

„… possible tokens, the tensors representing a level can become very high-dimensional and sparse. To deal with …” „… Voxel-basiertes Spiel. Da es in Minecraft eine große Anzahl möglicher Tokengibt, können die Tensoren, die ein …”

„… tensors.As the name suggests, the MSE is the mean of the squares of errorsbetween of target y and predicted …” „… output value for that input window. This way, theoutput tensor will be smaller than the input, depending …”

„… accordingto Tab. 4.1, any level can be converted into a tensor with only 9 tokenlayers instead of however many …” „… touse this kind of generator to create new, unseen level designs. Given thatlevel tensors in scale N+ 1 …”