Hello, I am looking for papers about the pros and cons of CNNs and RNNs, and the advantages of a hybrid CNN-RNN model over the two separate models (if indeed there is an advantage) in speech recognition tasks, or in event detection tasks. Can anyone suggest relevant studies?
Personally, I'd always go for CNNs for pattern recognition tasks such as in recognizing handwritten text, when I want to model sequential data (as in spoken language or for text-meaning classificytion tasks), RNNs usually work better. Up to now, I have never tried to combine both but here is an interesting paper that did that in your context:
Preprint ContextNet: Improving Convolutional Neural Networks for Auto...
Hey Annika Hamachers ,
Thank you for your reply.
There are some studies that used this Hybrid technique.
It makes sense that implementing CNN layers to extract features from spectrograms of audio files and then feeding them to RNN layers would work well, especially when you don't have a lot of data.
I am just wondering if someone made comparisons or discussed the advantages or disadvantages of each model.
Article Fake news detection: A hybrid CNN-RNN based deep learning approach
Thank you for your reply Aravinda C V , I think that I have a look at it
- Badges
- Science topic
- Similar topics
- Papers