Neural networks and data compression

Question

Neural networks

I have been interested in compression coding for a long time. And since I am currently working on neural networks and genetic algorithms, the question arose, what are the prospects for their use for compression.

Found a few on the internet. For example:
- M. Mahoney "Fast Text Compression with Neural Networks" ,
- P. Watta et al. "Image Compression using Backprop" ,
- Image Compression with Neural Networks ,
- SA Durai "Image Compression with Back-Propagation Neural Network using Cumulative Distribution Function" ,
- SAK Jilani "JPEG Image Compression using FPGA with Artificial Neural Networks" ,
-F. Ibrahim "Image Compression using Multilayer Feed Forward Artificial Neural Network and DCT" .

(The last two publications date back to 2010, but this is rather an exception. The works are usually old. It is, in general, understandable, people usually publish in paper journals.)

But these works are devoted to some kind of private research. I would like to see a general overview.

How promising is the use of ANNs (the mentioned Mahoney quite successfully applied them in his PAQ starting from version 7), what are the current tasks, what are the directions of research?

Answer 1

There are quite a lot of works on compression using ANN . There is also a general overview (only dated 1995).
Then everything is simple: click on "Cited by 159" and see who links to this review - it will be both individual articles and new reviews on this and related topics. Read, then see who referred to these articles - and so on in depth. Start with well-cited papers; if no one links to an article, it is either new or not particularly useful.
As for the prospects, it is better to ask the supervisor. As for the current tasks / directions - it will become clear from the articles.