Medical image classification

[BoltzmannEntropy]

Hi,
I wanted to inquire about the availability of the code for training models in Paddle-quantum, specifically related to medical image classification as found in this link: https://github.com/PaddlePaddle/Quantum/blob/master/applications/medical_image_classification/introduction_en.ipynb.

I was wondering if Paddle-quantum has a similar code implementation to Qiskit's quantum convolutional neural network as seen in this link: https://github.com/Qiskit/qiskit-machine-learning/blob/main/docs/tutorials/11_quantum_convolutional_neural_networks.ipynb.

Additionally, I am curious if it's possible to utilize quantum encoding for 256x256 medical images using Paddle-quantum on a 12GB GPU.?

Assuming I extract features using a classical CNN such as VGG12 and create a classical feature vector of 128 features, what would be the best quantum encoding method to use in Paddle-quantum? How many qubits would I need?

Thanks!

[gxli2017]

The problem of data encoding is currently a highly challenging issue, with common methods including amplitude encoding, angle encoding, and so on. You can refer to the following for useful info
https://qml.baidu.com/tutorials/machine-learning/encoding-classical-data-into-quantum-states.html

Hope it can help you！