In order to solve the problem of feature information loss caused by the introduction of a large number of pooling layers in traditional convolutional neural networks, based on the feature of Capsule Network (CapsNet)——using vector neurons to save feature space information, a network model 3DSPNCapsNet (3D Small Pooling No dense Capsule Network) was proposed for recognizing 3D models. Using the new network structure, more representative features were extracted while the model complexity was reduced. And based on Dynamic Routing (DR) algorithm, Dynamic Routing-based algorithm with Length information (DRL) algorithm was proposed to optimize the iterative calculation process of capsule weights. Experimental results on ModelNet10 show that compared with 3DCapsNet (3D Capsule Network) and VoxNet, the proposed network achieves better recognition results, and has the average recognition accuracy on the original test set reached 95%. At the same time, the recognition ability of the network for the rotation 3D models was verified. After the rotation training set is appropriately extended, the average recognition rate of the proposed network for rotation models of different angles reaches 81%. The experimental results show that 3DSPNCapsNet has a good ability to recognize 3D models and their rotations.
The way of selecting evolutionary parameters is vital for the optimal performance of the Quantum-inspired Evolutionary Algorithm (QEA). However, in conventional QEA, all individuals employ the same evolutionary parameters to complete update without considering the individual difference of the population, thus the drawbacks including slow convergence speed and being easy to fall into local optimal solution are exposed in computing combination optimization problem. To address those problems, an adaptive evolutionary mechanism was employed to adjust the rotation angle step and the quantum mutation probability in the quantum evolutionary algorithm. In the algorithm, the evolutionary parameters in each individual and each evolution generation were determined by the individual fitness to ensure that as many evolutionary individuals as possible could evolve to the optimal solution direction. In addition, the adaptive-evolution-based evolutionary algorithm needs to evaluate the fitness of each individual, which leads to a longer operation time. To solve this problem, the proposed adaptive quantum-inspired evolutionary algorithm was parallel implemented in different universe to improve the execution efficiency. The proposed algorithms were tested by searching the optimal solutions of three multimodal functions and solving knapsack problem. The experimental results show that, compared with conventional QEA, the proposed algorithms can achieve better performances in convergence speed and searching the global optimal solution.