Implement Quanvolutional Layer structure

[SaashaJoshi]

Requires implementation of the Quanvolutional Neural Network Layer structure.

Reference:

M. Henderson, S. Shakya, S. Pradhan, and T. Cook, “Quanvolutional Neural Networks: Powering Image Recognition with Quantum Circuits,” arXiv:1904.04767 [quant-ph], Apr. 2019, Available: https://arxiv.org/abs/1904.04767

[SaashaJoshi]

Quanvolutional structure is rather implemented as a layer in existing QCNN structures. It should, hence, directly inherit form the BaseLayer class.

[SaashaJoshi]

There also might be a need to build a QuanvolutionalNeuralNetwork class that implements the sequence abstract method and deals with measurement and results at the end of a QuanvolutionalLayer. The results received after the implementation of this layer are either processed classically (via some CNN) or need to be embedded again into a quantum circuit and processed with some QCNN structure.

[SaashaJoshi]

Is the quanvolutional layer always applied as the first layer in the neural network structure? Why cannot it be applied in between, lets say, convolutional and pooling layer?

[sharpaxis]

hey, could you briefly tell me more about this issue, I'd be interested In working on it

[SaashaJoshi]

That's awesome, @sharpaxis !

[SaashaJoshi]

Hi @sharpaxis, any updates? Would you need help anywhere?