With multiple graph pooling layers, the input graphs are hierachically coarsened to one node.
With some advanced algorithms, the new technologies are expected to control the production quality based on the digital images.
Next, a novel graph neural network with an attention mechanism is designed to map each subgraph into an embedding vector.
To design more efficient lightweight concolutional neural netwok, Depthwise-Pointwise-Depthwise inverted bottleneck block (DPD block) is proposed and DPDNet is designed by stacking DPD block.
We propose a taxonomy in terms of three levels, i. e.~structure level, algorithm level, and implementation level, for acceleration methods.