最新的Huawei HCIP-AI-EI Developer V2.5 - H13-321_V2.5免費考試真題

A. Object detection

B. Video analytics

C. Image classification

D. Speech recognition

A. SSD

B. R-CNN

C. YOLO

D. Faster-R-CNN

A. A transformer model directly captures the dependency between different positions in the input sequence through the self-attention mechanism, without using the recurrent neural network (RNN) or convolutional neural network (CNN).

B. Transformer models outperform RNN and CNN in processing long texts because they can effectively capture global dependencies.

C. Multi-head attention is the core component of a transformer model. It computes multiple attention heads in parallel to capture semantic information in different subspaces.

D. Positional encoding is optional in a transformer model because the self-attention mechanism can naturally process the order information of sequences.

A. Color image input is multi-channel, whereas text input is single-channel.

B. For CNN, there is no difference in handling text or image tasks.

C. When the CNN is used for text tasks, the kernel size must be the same as the number of word vector dimensions. This constraint, however, does not apply to image tasks.

D. CNNs are suitable for image tasks, but they perform poorly in text tasks.

A. Sigmoid

B. Tanh

C. SeLU

D. ReLU

A. FALSE

B. TRUE

A. Recursion-based method

B. Rule-based method

C. Deep learning-based method

D. Statistics-based method