Project Idea proposal for contribution to GSoC 2020

Hi! I am a GSoC 2020 aspirant, looking forward to contribute further to Model Zoo and Flux.Till now I have worked on following PRs:
Model zoo:
#192
#194
#196
Flux:
https://github.com/FluxML/Flux.jl/pull/1015
https://github.com/FluxML/Flux.jl/pull/1019
And I am very excited to contribute further.
I have few ideas for my proposal and would like to check their feasibility. I have planned to work on following two ideas :
1. Automatic Speech Recognition: In this, I plan to implement models: Deep Speech(RNN based model) and Wav2Letter(CNN based model).
2. Text to Image Synthesis: In this, I am planning to work with some GAN models like StackGANs for text to image synthesis.

Other than that I am also thinking over the idea of implementing models for Face Recognition like VGGFace as they haven't been implemented yet and about adding Tutorial blogs on the models implemented in Julia till now. I have prior experience working with Text Processing models, and Computer Vision models using Deep Learning. I have also worked with Generative Adversarial Networks in the past. Furthermore, I am open to other suggestions as well.

Thank you
@dhairyagandhi96 @MikeInnes @ViralBShah @avik-pal

The proposal mentions Model Zoo and prior work in PRs #192, #194, #196, plus Flux PRs #1015 and #1019, but names no files or tests. Review those PRs first, then seek maintainer guidance to narrow the speech-recognition, text-to-image, face-recognition, or tutorial ideas; completion criteria are not defined here.

Written by the indexing model from the issue text.